NAFLD’s Predisposion: insight from phenome-wide association and Mendelian Randomization

NAFLD’s Predisposion: insight from phenome-wide association and Mendelian Randomization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article NAFLD’s Predisposion: insight from phenome-wide association and Mendelian Randomization Huanyu Jiang, Qian Dai, Haiying Yan, Quanyu Du, Tao Shen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4722888/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Genome-wide association studies (GWAS) meta-analysis have unveiled common single nucleotide polymorphisms (SNPs) associated with the increased risk of Non-alcoholic fatty liver disease (NAFLD). We conducted a phenome-wide association study (PheWAS) using data from the UK Biobank to further elucidate NAFLD-associated phenotypes and investigate the disease's underlying biology. A significant enrichment was found in 31 of 778 traits examined using 17 known NAFLD-risk SNPs along with 4:1 matched control SNPs. We explored genetic correlations and causal relationships by employing bidirectional Mendelian randomization (MR) and linkage disequilibrium score regression (LDSC). Notably, strong positive genetic correlations with NAFLD were observed for BMI (r g = 0.73), Trunk fat mass (r g = 0.67), Type 2 diabetes (r g = 0.86), and weight (r g = 0.57), while whole-body impedance (r g = -0.31) and neutrophil count (r g = -0.28) exhibited negative correlations. Our MR analysis demonstrated unidirectional effects of BMI (OR = 1.57), Trunk fat mass (OR = 1.40), Weight (OR = 1.22), whole-body Impedance (OR = 0.83), and Type 2 diabetes (OR = 1.42) on NAFLD risk. Intriguingly, bidirectional causal effects were identified between Alcohol intake frequency and NAFLD (OR Alcohol intake frequency → NAFLD = 1.42; OR NAFLD → Alcohol intake frequency = 1.02), suggesting a complex interplay. Furthermore, through intermediary MR analyses, we uncovered pathways mediated by FGF21 and IL-10RB, linking BMI and Trunk fat mass, respectively, to NAFLD development. These findings provide novel insights into the multifaceted genetic landscape of NAFLD, highlighting the importance of body composition, metabolic health, and lifestyle factors in its pathogenesis. nonalcoholic fatty liver disease mendelian randomization phenome-wide association study linkage disequilibrium score regression Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Non-alcoholic fatty liver disease (NAFLD) is a growing health problem worldwide, characterized by the accumulation of fat in liver cells even in individuals who consume little or no alcohol 1 . It is a multifactorial disease in which genetic, environmental and lifestyle factors combine to contribute to its pathogenesis and progression 2 . An in-depth understanding of the complex relationships between NAFLD and its associated phenotypes is essential for the development of targeted interventions and treatments. Recent advances in the fields of genomics and epidemiology have provided new tools to analyze these relationships. An GWAS meta-analysis found 17 genetic risk alleles shared at 17 statistically independent loci that were associated with NAFLD 3 . However, a limited number of functional genomics studies reveal the biological consequences associated with single nucleotide polymorphisms (SNPs) through in vitro and in vivo analyses. In addition to traditional GWAS approaches, the use of polygenic score, Mendelian randomization (MR), gene pathway analyses, and PheWAS can expand our understanding of NAFLD pathogenesis. Phenome-Wide Association Studies (PheWAS) are strategically devised to explore across a multitude of phenotypes for identifying associations with genetic variants 4 . This methodology illuminates the potential influences of these variants on an extensive array of diseases and traits. The implementation of PheWAS has been significantly advanced by the advent of large-scale biobanks and genomic consortia, which facilitate comprehensive screenings across a wide range of clinical outcomes. By expanding the range of phenotypic expressions linked to a particular genetic alteration, performing Phenome-Wide Association Studies (PheWAS) on a Single Nucleotide Polymorphism (SNP) known to be related to a certain illness can uncover possible concurrent disorders. It can also identify characteristics that could play a role in the linkage between the SNP and the illness in question. For instance, a data-driven PheWAS study conducted within the UK Biobank explored the correlation between elevated Body Mass Index (BMI) and various disease outcomes 5 . This study utilized genetic risk scores and assessed causal relationships through MR analysis. It highlighted the multifaceted impact of BMI on numerous phenotypes, underscoring the utility of PheWAS in understanding complex traits. However, current research has not yet leveraged PheWAS to identify other "intermediate traits" that connect NAFLD risk alleles with the causal mechanisms underlying NAFLD pathogenesis. Therefore, the integration of PheWAS and MR presents a promising approach in NAFLD research, offering a deeper understanding of the genetic and phenotypic framework linked to NAFLD. This approach provides valuable information regarding its etiology, progression, and potential therapeutic targets. Our study aims to systematically explore the genetic basis of NAFLD and its wide-ranging phenotypic consequences using these methodologies. Methods The datasets used were summary data; all informed consent and ethical approval were obtained in the original studies. Selection of NAFLD risk alleles We compiled a selection of genetic variants frequently connected with NAFLD, ensuring they had a minor allele frequency of at least 1% among populations of European descent, to be examined in our PheWAS study. The selection drew upon the latest and most extensive genome-wide association study (GWAS) meta-analysis focused on NAFLD at that point, which integrated data from four cohorts with electronic health records confirming NAFLD in individuals of European background (comprising 3,584 cases and 621,081 controls) 3 . For the PheWAS inquiries, we progressed with 17 independent variants linked to NAFLD. This set encompassed 11 SNPs initially pinpointed through GWAS of NAFLD, along with an additional 6 new variants whose associations were validated in that same report. Control SNP set Prior studies have indicated that beyond a ratio of one case for every four controls, the gain in statistical power tends to be negligible 6 . Threrefore, for the 17 NAFLD risk SNPs, we generated a set of control SNPs at a 1:4 ratio using SNPsnap from the 1000 Genomes Project. We meticulously chose four control SNPs to correspond with those associated with NAFLD, adhering to multiple matching parameters: minor allele frequency within a 5% margin, gene density in the vicinity within a 50% range, proximity to the closest gene also within a 50% range, and to approximate the extent of the haplotype block, the count of surrounding SNPs in linkage disequilibrium (LD) with an R2 of 0.50 or above, allowing for a 50% variance. PheWAS analyses GeneATLAS database compiles GWAS data for 118 quantitative and 660 binary traits. The database catalogs associations with 9,113,133 genetic variants—both genotyped and imputed—from 452,264 UK Biobank participants of European ancestry. It supports queries for genetic or phenotypic data to assess genotype-phenotype associations. In this study, we used the GeneATLAS to identify associations between traits and established NAFLD risk SNPs, along with a control set of 68 SNPs identified through SNPsnap. For each queried variant, we extracted summary statistics related to associated traits from GeneATLAS for subsequent analysis. Following the methodology outlined by Semmes et al., SNP-trait linkages that showed a nominal level of significance ( p -value less than 0.01) were advanced for further trait-enrichment analysis 7 . This process was designed to compare traits associated with NAFLD-linked SNPs against those linked with the control SNP set. Given the exploratory nature of PheWAS, the significance threshold for SNP-trait associations was determined with a degree of arbitrariness, provided it remained consistent across all analyzed SNP sets. This approach allowed for a uniform standard for comparison and analysis, yet with the flexibility to accommodate the intrinsic variability and genetic correlations among multiple traits studied. For instance, considering the high genetic correlation between traits such as weight and BMI, or reticulocyte percentage and reticulocyte count, employing a more stringent p- value threshold could potentially be overly conservative. Such a strict threshold might overlook SNP-trait associations that, despite not meeting the highest levels of statistical stringency, could still hold biological relevance and significance in the context of NAFLD. This rationale underscores the importance of balancing statistical rigor with the practical considerations of genetic correlation and trait complexity in PheWAS analyses. PheWAS analyses were conducted to explore the associations between 17 NAFLD-specific SNPs and a broad spectrum of 778 phenotypic traits. These associations were then benchmarked against the outcomes of a PheWAS analysis for a control set of SNPs, utilizing R version 4.1.3 for all statistical computations. To ascertain whether specific traits were enriched in association with the NAFLD risk variants compared to the control SNP set, we employed Fisher's exact test. This approach enabled us to compare how often specific traits occurred in association with SNPs related to NAFLD as well as with the control SNPs, aiming to identify significant trait enrichments linked to the known NAFLD risk variants. To account for the challenge of multiple comparisons inherent in such a broad analysis, the p- values obtained from Fisher's exact test for trait enrichment were subjected to Bonferroni correction. This adjustment ensures that the significance of our findings reflects a balanced consideration of both the potential for genuine associations and the statistical risk of type I errors due to the extensive number of tests performed. Linkage disequilibrium score (LDSC) analysis The LDSC method serves as a powerful analytical tool for unraveling the genetic underpinnings of complex human traits. Utilizing summary-level data from GWAS, LDSC enables the quantification of heritability attributable to common genetic variants for diseases and facilitates the assessment of genetic overlap between traits. Our research concentrated on calculating the genetic correlation (r g ) among NAFLD and related phenotypic traits, shedding light on the mutual genetic foundation that could be influencing these conditions. Further, we assessed the heritability (h 2 ) of these phenotypes, which reflects the proportion of phenotypic variance that can be attributed to genetic factors. This analysis extended to computing the genetic correlations among the phenotypes themselves, aiming to identify potential genetic bridges that connect various traits within the context of NAFLD. To ensure the robustness of our findings, we applied the False Discovery Rate (FDR) correction to the p- values obtained for both the r g and h 2 estimates. The developers of the method have made these scores accessible through GitHub (https://github.com/bulik/ldsc). Two-sample mendelian randomization (TSMR) analysis Bidirectional TSMR analyses were performed to investigate the potential causative links between NAFLD and the phenotypes pinpointed via PheWAS, adhering to the STROBE-MR recommendations. The GWAS summary statistics for NAFLD were sourced from an electronic health record-based GWAS meta-analysis conducted by Nooshin Ghodsian and colleagues 8 , with data on other associated phenotypes detailed in Supplementary Table 1. In our study, we employed distinct GWAS meta-analysis data sources for the selection of NAFLD risk alleles and the TSMR analysis. This strategic approach was driven by our objectives for each analysis phase. For the Phewas analysis, our goal was to incorporate as many risk alleles as possible to comprehensively identify potential associations with NAFLD, leveraging a wider array of genetic variants. Conversely, for the TSMR process, the emphasis was on maximizing the total sample size and the number of SNPs to enhance the statistical power and robustness of our causal inference analysis. In this study, we primarily utilized the Generalized Summary-data-based Mendelian Randomization (GSMR) method in Genome-wide Complex Trait Analysis (GCTA) software v1.94.1. This choice was motivated by GSMR's capability not only to estimate the causal effects of exposure factors on outcomes but also to identify and mitigate the effects of pleiotropy 9 . We chose near-independent SNPs from GWAS data using a linkage disequilibrium (LD) threshold of r² = 0.05 within a 1 Mb region and a significance threshold of p < 5.0 × 10^−8. The selection was based on LD estimates from the 1000G project's European population dataset, employing PLINK v2.0.0 alpha. In instances where a phenotype had fewer than 10 independent lead SNPs meeting the rigorous GWAS significance level, we lowered the threshold to p < 1.0 × 10^−5 to maintain an adequate pool of SNPs for our analysis. To mitigate the impact of horizontal pleiotropy on our results, we used heterogeneity in dependent instrument (HEIDI)-outlier detection as part of the GSMR method. This step, taken before the analysis, involved applying the default HEIDI-outlier significance threshold of 0.01 to filter out unsuitable SNPs. Additionally, to assess the robustness of our results, we employed several other MR methods, including Inverse Variance Weighted (IVW), Weighted Median, and Weighted mode estimation with the R package “TwoSampleMR”. Furthermore, we performed MR-Egger regression to identify instances of horizontal pleiotropy, and MR-PRESSO was used to adjust for bias introduced by pleiotropic genetic variants. In situations where the GWAS data sources did not meet the computational requirements for GSMR in reverse MR analyses, alternative methods were utilized. Mediation analysis In our research, we utilized a two-sample Mendelian randomization strategy to pinpoint phenotypes with a causal link to NAFLD. The aim was to delve deeper into the mediators through which these phenotypes precipitate the onset of NAFLD. Central to this investigation are inflammatory mediators, which act as a critical bridge connecting various metabolic and immune pathways associated with NAFLD. These mediators orchestrate a series of mechanisms, including insulin resistance 10 , oxidative stress 11 , and the activation of the innate immune system 12 . Their actions contribute significantly to the progression of liver damage, evolving from simple steatosis to more severe conditions like Non-Alcoholic Steatohepatitis (NASH), fibrosis, and, in some cases, cirrhosis 13 . Utilizing data from a study that explored genome variant associations with 91 plasma proteins across 14,824 European participants, it is possible to uncover potential intermediary variables that mediate various phenotypes leading to NAFLD 14 . Our initial step involved using TSMR analysis to identify factors causally related to NAFLD, focusing on inflammatory proteins as exposure factors. Next, we treated the phenotypes causally linked to NAFLD as exposure factors and the inflammatory proteins, which were established as causally related to NAFLD, as outcome variables. In cases where exposures had a significant impact on outcomes, the 'coefficient product' approach was implemented to evaluate the indirect effects exerted by these phenotypes on NAFLD via mediating pathways. This method involves calculating the mediator's effect by multiplying the influence of the exposure on the mediator with the mediator's impact on the outcome. To understand the contribution of each mediator to the overall outcome, we calculated the proportion of the mediating effect relative to the total effect of the exposure on the outcome using the formula (β1 × β2/β3). Results Figure 1 : Research Framework. This figure delineates our methodology for exploring associations between phenotypes and established disease risk variants, aiming to pinpoint phenotypes for further LDSC and MR analysis. When applied to NAFLD, this strategy led to the identification of 31 phenotypes significantly associated with NAFLD. It also facilitated subsequent median analyses to assess how variants linked to inflammatory proteins contribute to the risk of NAFLD. Table 1 SNPs associated with NAFLD, validated in previous GWAS meta-analyses, and incorporated into our enrichment analysis within the PheWAS framework. Position (Chr:BP) a rsID MAF Function b Nearest gene 22:44324730 rs738408 0.23 exonic PNPLA3 19:19379549 rs58542926 0.07 exonic TM6SF2 19:45411941 rs429358 0.16 missense APOE 2:27730940 rs1260326 0.41 exonic GCKR 8:126506694 rs112875651 0.38 intron variant TRIB1 10:113947040 rs4918722 0.31 intron variant GPAM 1:220970028 rs2642438 0.28 exonic MARC1 4:100505326 rs7661964 0.26 3’ UTR MTTP 9:132566666 rs7029757 0.11 exonic TOR1B 4:100239319 rs1229984 0.03 exonic ADH1B 16:53813367 rs17817449 0.41 intron variant FTO 2:165555539 rs79953491 0.15 intron variant COBLL1 19:7218635 rs112630404 0.16 intron variant INSR 19:54677001 rs626283 0.44 intergenic TMC4/MBOAT7 17:17979099 rs4561528 0.37 NA SREBF1 9:10462423 rs10756038 0.29 intron variant PTPRD 11:823586 rs140201358 0.01 exonic PNPLA2 a Position in GRCh37/hg19. b Minor allele frequency in European-ancestry individuals from the 1000 genomes project. Phewas analysis PheWAS were conducted utilizing the UK Biobank GeneATLAS database to assess the association of each NAFLD-linked variant and control variants, with a comprehensive array of 778 traits. These traits could be organized into 24 distinct categories. Among the 68 control SNPs evaluated, six variants (rs186719489, rs111571790, rs139635278, rs962745, rs118065116, rs61952239) were not present in the GeneATLAS database. Consequently, our analysis proceeded with 17 test SNPs and 62 control SNPs included in the enrichment analysis. Within the GeneATLAS database, we found that 329/778 traits exhibited a nominal association ( p < 0.01) with at least one of the 17 SNPs that have been linked to NAFLD. These traits were further subjected to enrichment analysis against the control SNP set, as detailed in Supplementary Table S2. Out of the traits assessed, 128 were determined to have a higher probability of association with the SNPs linked to NAFLD than with the control SNPs ( p < 0.05). Following Bonferroni correction, 31 of these traits retained their significance, as illustrated in Fig. 2 . Notably, the category with the highest number of significantly enriched traits was physical and body measurements, encompassing 10 distinct phenotypes, followed by genetic and biochemical markers. LDSC analysis From the 31 phenotypes significantly associated with NAFLD, we excluded E10-E14 Diabetes mellitus due to unclear definitions. Instead, we incorporated Type 1 and Type 2 diabetes for further analysis. K76 Other diseases of liver, due to overlapping diagnostic codes with NAFLD, E78 Disorders of lipoprotein metabolism and other lipidaemias, I20-I25 Ischaemic heart diseases and E11 Non-insulin-dependent diabetes mellitus, all of which had no related GWAS data available, were also excluded. Consequently, 27 traits were included in the subsequent analysis. Initially, we conducted pairwise LDSC analysis on the selected 27 traits, identifying several pairs with significant genetic correlations, as depicted in Fig. 3 . Notably, there were pronounced genetic correlations between certain physical and body measurements and metabolic diseases. Specifically, Type 2 diabetes, Arm fat-free mass, and BMI exhibited substantial positive r g . Conversely, Impedance of the whole body demonstrated negative genetic correlations with the aforementioned indicators. Subsequently, we analyzed the heritability of these 27 traits as well as their genetic correlation with NAFLD. As depicted in Fig. 4 A, the LDSC analysis revealed a strong positive correlation of NAFLD with several key traits: BMI showing a genetic correlation of 0.73, Trunk fat mass with r g = 0.67, Type 2 diabetes with r g = 0.86, and weight with r g = 0.57. Additionally, a negative correlation of NAFLD with Impedance of the whole body (r g = -0.31) and Neutrophil count (r g = -0.28) was also observed. Figure 4 B highlights the traits with the highest heritability estimates, including Whole body fat-free mass (h 2 = 0.31), Trunk fat-free mass (h 2 = 0.31), and Trunk predicted mass (h 2 = 0.31). TSMR analysis We further investigated bidirectional causal associations between 27 traits and NAFLD. Our ability to explore causal associations using GSMR, with NAFLD as the exposure and other traits as outcomes, was limited due to an insufficient number of lead SNPs for NAFLD (n < 10 at p < 1.0 × 10^−5). Consequently, in the reverse MR analysis, we employed five alternative methods, with the IVW method serving as the primary analysis. As illustrated in Fig. 5 A-B and Supplementary table 4, we identified unidirectional progressive effects of BMI, Trunk fat mass, Weight, and Type 2 diabetes on NAFLD. Conversely, Impedance of the whole body was identified as a protective factor against NAFLD. The GSMR results for each of these phenotypes were corroborated by at least two other methods. Furthermore, we observed strong bidirectional progressive effects of Alcohol intake frequency on NAFLD, indicating a reciprocal influence between alcohol consumption and NAFLD risk. This bidirectional MR analysis was validated in at least two methods besides the primary analysis, reinforcing the complex interplay between alcohol intake and NAFLD (Supplementary Table 5). Median analysis Utilizing the 'product of coefficients' approach, we pinpointed 2 potential causal routes that could influence the development of NAFLD (Fig. 6 and Table S6). The effect of BMI on NAFLD was partially mediated by Fibroblast growth factor 21 (FGF21) (indirect effects = 0.030, mediated proportion: 6.48%). Trunk fat mass might increase the risk of NAFLD by increasing Interleukin-10 receptor subunit beta (IL-10RB) (indirect effects = 0.014, mediated proportion: 4.00%). Discussion We adopted an innovative and corroborated model to sift through extant GWAS and PheWAS datasets, aiming to identify characteristics correlated with NAFLD susceptibility. Initially, we curated SNPs linked to NAFLD from extensive comprehensive GWAS meta-analysis and paired these risk variants to control SNPs. Through PheWAS conducted on these SNPs utilizing UK Biobank data, 31 traits were identified as significantly enriched for associations with NAFLD risk SNPs. LDSC further enabled us to determine the genetic correlations among these 31 traits, unveiling both positive and negative genetic correlations with NAFLD. TSMR analysis, with GSMR as the primary method, implicated six traits as playing a causal role in NAFLD pathogenesis. PheWAS results indicated that under the category of physical and body measurements, ten traits were significantly enriched for associations with NAFLD. However, only BMI, Trunk fat mass, and weight were suggested by TSMR analyses to exhibit vertical pleiotropy, directly contributing to NAFLD risk. Elevated levels of these indicators are indicative of obesity, a condition prevalent in up to 80% of NAFLD patients. Obesity is considered as a state of chronic low-grade inflammation and is associated with various complications including NAFLD. The mechanisms through which obesity leads to NAFLD involve complex interactions among metabolic processes, including insulin resistance, altered lipid metabolism, and inflammation. Specifically, obesity-induced insulin resistance is a central factor that exacerbates hepatic fat accumulation through increased hepatic fatty acid influx and lipogenesis, coupled with decreased fatty acid oxidation 15 . Additionally, obesity leads to elevated levels of free fatty acids released by adipose tissue, which further augment hepatic uptake and synthesis of lipids, potentially reducing β-oxidation and contributing to hepatic fat accumulation 16 . Inflammatory factors, notably TNF-α, IL-1β and IL-6 produced especially by visceral fat, can promote subacute hepatic inflammation and fibrosis through various signaling pathways 17 . Moreover, obesity alters the secretion profile of hormones and cytokines by adipocytes, such as reduced adiponectin and increased leptin levels, impacting overall metabolic state and liver metabolism 18 . Our study highlights the nuanced relationship between different physical measurement indicators related to body composition and NAFLD. While BMI, Trunk fat mass, and Weight were identified as causal factors for NAFLD, other body physical measurements such as Left arm fat-free/predicted mass, Right arm fat-free/predicted mass, and Hip circumference, despite their association with NAFLD in PheWAS results and positive genetic correlations as shown by LDSC, did not emerge as causal in TSMR analyses. This discrepancy underscores the complexity of NAFLD's etiology, suggesting that not all obesity-related markers carry the same risk for NAFLD. Specifically, our findings align with the notion that central obesity, as reflected by Trunk fat mass, plays a more pivotal role in NAFLD pathogenesis compared to other fat distribution patterns. For instance, studies have shown that visceral fat, rather than subcutaneous fat, is more metabolically active and contributes significantly to hepatic steatosis and insulin resistance, thereby increasing the risk of NAFLD 19 . Furthermore, the differential impact of fat depots on NAFLD risk underscores the importance of targeted interventions that reduce visceral fat to mitigate NAFLD risk, beyond the general reduction of body weight or BMI. Our study, therefore, emphasizes the significance of considering specific fat distribution patterns in NAFLD risk assessment and management strategies. Whole-body impedance is a valuable tool for assessing body composition, including distinctions between fat and lean tissues. It is generally observed that higher whole-body impedance values correlate with a lower amount of body fat 20 . This relationship underpins our finding that whole-body impedance acts as a protective factor against NAFLD. In the context of clinical outcomes, the Fat-Free Mass Index (FFMI), derived from whole-body impedance measurements, shows promising utility. It has demonstrated comparable, and in some cases superior, predictive power for NAFLD risk compared to traditional metrics such as BMI 21 . This suggests that incorporating whole-body impedance assessments into clinical practice could provide a more nuanced understanding of NAFLD risk, beyond conventional obesity measurements . Notably, the association between bioelectrical impedance-derived metrics and liver health has been observed in various studies. For instance, research has shown that bioelectrical impedance analysis can effectively reflect changes in body composition, including muscle mass, which is inversely associated with the severity of hepatic steatosis in patients with NAFLD 22 . This underscores the potential of whole-body impedance measurements not only as a diagnostic tool but also as a predictor of NAFLD progression, highlighting the importance of lean mass preservation in NAFLD management strategies. Our findings highlight a bidirectional causal relationship between alcohol intake frequency and NAFLD. The effects of Alcohol intake frequency on NAFLD (OR = 1.42) were stronger than those of NAFLD on Alcohol intake frequency (OR = 1.02). The odds ratio of 1.42 suggests that with each ascending level of alcohol intake frequency—from 'Never,' 'Special occasions only,' '1–3 times a month,' 'Once or twice a week,' 'Three or four times a week,' to 'Daily or almost daily'—the risk of developing NAFLD increases by 42%. Conversely, the presence of NAFLD is associated with a slight increase in the frequency of alcohol consumption. However, this increase is relatively modest when compared to the impact of alcohol intake frequency on NAFLD risk. Traditionally, heavy drinking has been unequivocally linked to liver damage, while the effects of non-heavy consumption remain contentious 23 . A notable prospective cohort study suggested that even moderate drinking might exacerbate fibrosis in NAFLD patients, underscoring potential risks of any alcohol intake in this population 24 . Contrarily, some cross-sectional studies report no adverse effects from moderate alcohol use on NAFLD, though these findings could be skewed by factors like temporal uncertainty and reverse causality, where sicker patients might abstain from drinking 25 . FGF21, a hormone primarily expressed in the liver, plays a pivotal role in regulating lipid and glucose metabolism, boasting capabilities to enhance insulin sensitivity and promote fatty acid oxidation. Moreover, FGF21 exhibits potential in inhibiting the development of NAFLD through its promotion of fatty acid oxidation and its ability to reduce hepatic fat accumulation via insulin-independent pathways 26 . Despite the therapeutic potential of FGF21 in treating metabolic diseases, elevated levels of FGF21 in individuals with obesity and NAFLD suggest a possible dysfunction or resistance to FGF21 signaling in these conditions 27 . Our study found that FGF21 acts as a mediating factor in the pathway from BMI to NAFLD, further emphasizing the significance of FGF21 in metabolic health. This indicates that an upregulation of FGF21 expression, in response to increased body weight, might be the body's attempt to cope with metabolic stress induced by obesity through enhancing lipid oxidation and improving metabolic health. However, the rise in FGF21 levels may also reflect a failure of metabolic adaptation. Additionally, through mediation MR analysis, IL-10RB was demonstrated to mediate the pathway through which trunk fat mass leads to NAFLD. IL-10RB, an integral part of the IL-10 signaling cascade, is instrumental in regulating immune responses and suppressing inflammation in adipose tissue and insulin resistance 28 . Recent studies have shown that IL-10 can directly inhibit the thermogenesis of adipocytes through a STAT3-dependent signaling pathway, and germline deletion of IL-10 can protect mice from insulin resistance and diet-induced obesity (DIO) 29 . Furthermore, the specific deletion of IL-10 or Blimp-1 in Treg cells can improve insulin sensitivity and DIO, underscoring that the inhibitory effect of Treg cells through IL-10 secretion extends beyond interactions among immune cells to also include the suppression of the beiging process in non-immune cells like adipocytes. Based on this, we infer that in the context of obesity and NAFLD, excessive fat accumulation, especially in the trunk, is closely associated with an increased inflammatory state. The overexpression of IL-10 under chronic inflammatory levels affects the functionality of adipocytes, disrupting the inflammatory balance and metabolic stability of adipose tissue, thereby promoting the development of NAFLD. Our research has certain constraints and justifiable issues regarding the combined GWAS-PheWAS methodology we have adopted. Specifically, our enrichment analysis of NAFLD risk SNPs across 778 traits involved Bonferroni correction to adjust for multiple comparisons. While this rigorous correction method minimizes the risk of false positives, it may also reduce the sensitivity to detect traits potentially sharing genetic variations with NAFLD, potentially excluding phenotypes of relevance. Additionally, our study's analyses were exclusively based on traits available in the U.K. Biobank, which predominantly features a European population. Consequently, the findings may not universally apply across different ethnic groups. For instance, while our results identify obesity-related metrics such as BMI as risk factors for NAFLD, it is important to note that not all obese individuals develop NAFLD. More critically, NAFLD can also occur in individuals who are not obese. Although NAFLD in non-obese individuals has been observed across various ethnicities, including children and adults, it is reported more frequently in Asian populations, even when using strict, ethnicity-specific BMI criteria for obesity 30 . This suggests that conducting similar analyses in diverse ethnic backgrounds could yield varying insights, underscoring the importance of considering ethnic diversity in understanding NAFLD's genetic predispositions. Conclusions In summary, our integration of GWAS and PheWAS datasets has illuminated the multifaceted etiology of NAFLD, uncovering both horizontal and vertical pleiotropy across several traits. Importantly, our findings implicate FGF21 and IL-10RB as significant players in the pathogenesis of NAFLD. These insights pave the way for more in-depth mechanistic studies aimed at understanding the specific contributions of these traits to NAFLD pathogenesis and exploring their interconnected relationships. The knowledge gained holds promise for informing the development of more targeted prevention and treatment strategies for NAFLD. Declarations Funding This research and the associated article processing charges were funded by the Sichuan Provincial Science and Technology Plan Project of Science and Technology Bureau of Sichuan, grant number 2023JDRC0092. Author Contribution H.Y.J. and Q.D. contribute equally to this work. Conceptualization, H.Y.J. and H.Y.Y; methodology, Q.D.; software, H.Y.J.; formal analysis, H.Y.J. and Q.D.; writing—original draft preparation, H.Y.J. and Q.D.; writing—review and editing, Q.D.; visualization, H.Y.J.; supervision, T.S. All authors have read and agreed to the published version of the manuscript. Data Availability Data is provided within the manuscript or supplementary information files References Younossi, Z. M. et al. Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 64 , 73–84 (2016). Estes, C., Razavi, H., Loomba, R., Younossi, Z. & Sanyal, A. J. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology 67 , 123–133 (2018). Chen, Y. et al. Genome-wide association meta-analysis identifies 17 loci associated with nonalcoholic fatty liver disease. Nat Genet 55 , 1640–1650 (2023). Bastarache, L., Denny, J. C. & Roden, D. M. Phenome-Wide Association Studies. JAMA 327 , 75–76 (2022). Hyppönen, E., Mulugeta, A., Zhou, A. & Santhanakrishnan, V. K. A data-driven approach for studying the role of body mass in multiple diseases: a phenome-wide registry-based case-control study in the UK Biobank. Lancet Digit Health 1 , e116–e126 (2019). Hennessy, S., Bilker, W. B., Berlin, J. A. & Strom, B. L. Factors influencing the optimal control-to-case ratio in matched case-control studies. Am J Epidemiol 149 , 195–197 (1999). Semmes, E. C. et al. Leveraging Genome and Phenome-Wide Association Studies to Investigate Genetic Risk of Acute Lymphoblastic Leukemia. Cancer Epidemiol Biomarkers Prev 29 , 1606–1614 (2020). Ghodsian, N. et al. Electronic health record-based genome-wide meta-analysis provides insights on the genetic architecture of non-alcoholic fatty liver disease. Cell Rep Med 2 , 100437 (2021). Zhu, Z. et al. Causal associations between risk factors and common diseases inferred from GWAS summary data. Nat Commun 9 , 224 (2018). Asrih, M. & Jornayvaz, F. R. Inflammation as a potential link between nonalcoholic fatty liver disease and insulin resistance. J Endocrinol 218 , R25-36 (2013). Xu, D. et al. The Role of Nrf2 in Liver Disease: Novel Molecular Mechanisms and Therapeutic Approaches. Front Pharmacol 9 , 1428 (2018). Arrese, M., Cabrera, D., Kalergis, A. M. & Feldstein, A. E. Innate Immunity and Inflammation in NAFLD/NASH. Dig Dis Sci 61 , 1294–1303 (2016). Buzzetti, E., Pinzani, M. & Tsochatzis, E. A. The multiple-hit pathogenesis of non-alcoholic fatty liver disease (NAFLD). Metabolism 65 , 1038–1048 (2016). Zhao, J. H. et al. Genetics of circulating inflammatory proteins identifies drivers of immune-mediated disease risk and therapeutic targets. Nat Immunol 24 , 1540–1551 (2023). Milić, S., Lulić, D. & Štimac, D. Non-alcoholic fatty liver disease and obesity: biochemical, metabolic and clinical presentations. World J Gastroenterol 20 , 9330–9337 (2014). Polyzos, S. A., Kountouras, J. & Zavos, C. Nonalcoholic fatty liver disease: the pathogenetic roles of insulin resistance and adipocytokines. Curr Mol Med 9 , 299–314 (2009). Boden, G. et al. Free fatty acids produce insulin resistance and activate the proinflammatory nuclear factor-kappaB pathway in rat liver. Diabetes 54 , 3458–3465 (2005). Monsour, H. P., Frenette, C. T. & Wyne, K. Fatty liver: a link to cardiovascular disease--its natural history, pathogenesis, and treatment. Methodist Debakey Cardiovasc J 8 , 21–25 (2012). Fracanzani, A. L. et al. Liver and Cardiovascular Damage in Patients With Lean Nonalcoholic Fatty Liver Disease, and Association With Visceral Obesity. Clin Gastroenterol Hepatol 15 , 1604-1611.e1 (2017). Foster, K. R. & Lukaski, H. C. Whole-body impedance--what does it measure? Am J Clin Nutr 64 , 388S-396S (1996). Elia, M. Body composition by whole-body bioelectrical impedance and prediction of clinically relevant outcomes: overvalued or underused? Eur J Clin Nutr 67 Suppl 1 , S60-70 (2013). Yodoshi, T. et al. Impedance-based measures of muscle mass can be used to predict severity of hepatic steatosis in pediatric nonalcoholic fatty liver disease. Nutrition 91–92 , 111447 (2021). Liangpunsakul, S. & Chalasani, N. What should we recommend to our patients with NAFLD regarding alcohol use? Am J Gastroenterol 107 , 976–978 (2012). Chang, Y. et al. Nonheavy Drinking and Worsening of Noninvasive Fibrosis Markers in Nonalcoholic Fatty Liver Disease: A Cohort Study. Hepatology 69 , 64–75 (2019). Ajmera, V. H., Terrault, N. A. & Harrison, S. A. Is moderate alcohol use in nonalcoholic fatty liver disease good or bad? A critical review. Hepatology 65 , 2090–2099 (2017). Liu, J., Xu, Y., Hu, Y. & Wang, G. The role of fibroblast growth factor 21 in the pathogenesis of non-alcoholic fatty liver disease and implications for therapy. Metabolism 64 , 380–390 (2015). Inagaki, T. Research Perspectives on the Regulation and Physiological Functions of FGF21 and its Association with NAFLD. Front Endocrinol (Lausanne) 6 , 147 (2015). Ouyang, W. & O’Garra, A. IL-10 Family Cytokines IL-10 and IL-22: from Basic Science to Clinical Translation. Immunity 50 , 871–891 (2019). Rajbhandari, P. et al. IL-10 Signaling Remodels Adipose Chromatin Architecture to Limit Thermogenesis and Energy Expenditure. Cell 172 , 218-233.e17 (2018). Kim, D. & Kim, W. R. Nonobese Fatty Liver Disease. Clin Gastroenterol Hepatol 15 , 474–485 (2017). Additional Declarations No competing interests reported. Supplementary Files SupplementaryTables.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4722888","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":327789210,"identity":"beeca601-1f53-4f6e-9446-725dcec6035b","order_by":0,"name":"Huanyu Jiang","email":"","orcid":"","institution":"Chengdu University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Huanyu","middleName":"","lastName":"Jiang","suffix":""},{"id":327789211,"identity":"ccbe19e5-c3cf-4ab3-a14b-260ca4576888","order_by":1,"name":"Qian Dai","email":"","orcid":"","institution":"Hospital of Chengdu University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Qian","middleName":"","lastName":"Dai","suffix":""},{"id":327789212,"identity":"fc9abfc5-e27c-4b95-9603-1e4b2091b0d1","order_by":2,"name":"Haiying Yan","email":"","orcid":"","institution":"Fuling District Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Haiying","middleName":"","lastName":"Yan","suffix":""},{"id":327789213,"identity":"efc6dcc7-4d32-44e9-8119-60c195ed8faf","order_by":3,"name":"Quanyu Du","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYNACNgY5BoYEKOcAkVqMgVoYG0jSkthAtBa+G8nPHvOUHU7vb88xf/Azh0GO70YC4+cCPFokb6SZG/OcO5w748wbw8bebQzGkjcSmKVn4NFicCPBTJq37XDuBokcw2bGbQyJG24ksDHz4NWS/g2kJd0AqqWeCC05YFsSYFoSDAhpkTzzpkxyzrl0wxlnnhXO7N0mYTjzzMNmaXxa+I6nb5N4U2Ytz9+evOHDz2028nzHkw9+xqcFFAtMSAokgBgaP/i0MP7Ar2QUjIJRMApGOgAAtL5O8B3ZP1gAAAAASUVORK5CYII=","orcid":"","institution":"Hospital of Chengdu University of Traditional Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Quanyu","middleName":"","lastName":"Du","suffix":""},{"id":327789214,"identity":"f1100b12-62bf-40e6-93bd-afc30fd25312","order_by":4,"name":"Tao Shen","email":"","orcid":"","institution":"Chengdu University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Shen","suffix":""}],"badges":[],"createdAt":"2024-07-11 08:44:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4722888/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4722888/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":61860662,"identity":"8250ee00-a40c-4769-bc7e-f6846b6176c1","added_by":"auto","created_at":"2024-08-06 10:47:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":111473,"visible":true,"origin":"","legend":"\u003cp\u003eResearch Framework. This figure delineates our methodology for exploring associations between phenotypes and established disease risk variants, aiming to pinpoint phenotypes for further LDSC and MR analysis. When applied to NAFLD, this strategy led to the identification of 31 phenotypes significantly associated with NAFLD. It also facilitated subsequent median analyses to assess how variants linked to inflammatory proteins contribute to the risk of NAFLD.\u003c/p\u003e","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/45291c2c53c6b6ca7f64f40e.png"},{"id":61861151,"identity":"67aa5181-4afd-4b52-b982-5e50172691b3","added_by":"auto","created_at":"2024-08-06 10:55:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":726082,"visible":true,"origin":"","legend":"\u003cp\u003eManhattan plot for 17 NAFLD-associated SNPs and 62 control SNPs with all phenotypes included by category. The horizontal axis shows the phenotypes by category, and the vertical axis shows the -log10 transformed P values adjusted by Bonferroni correction.\u003c/p\u003e","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/86578273ca2a5f22550c1f0f.png"},{"id":61860663,"identity":"da24c603-b499-4e4b-b16a-2d536e3070be","added_by":"auto","created_at":"2024-08-06 10:47:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":499340,"visible":true,"origin":"","legend":"\u003cp\u003eGenetic correlations among 27 traits identified through Phenome-Wide Association Study (PheWAS) analysis and subsequent quality control measures. The heatmap is structured such that the lower half below the diagonal displays the genetic correlations between traits, illustrating the extent to which their genetic determinants are shared or distinct. The upper half above the diagonal showcases the \u003cem\u003ep\u003c/em\u003e-values from the LDSC analysis, providing statistical significance for the observed genetic correlations\u003c/p\u003e","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/78e2cecbf7b54eff072f659d.png"},{"id":61860669,"identity":"15e332f1-a10d-4b20-bfe5-3f47e7c88e14","added_by":"auto","created_at":"2024-08-06 10:47:57","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":846509,"visible":true,"origin":"","legend":"\u003cp\u003eLDSC results between NAFLD and 27 traits associated with it. A. The outer cycle of the figure displays the genetic correlations between each of the 27 NAFLD-associated traits and NAFLD. The inner cycle presents the \u003cem\u003ep-\u003c/em\u003evalues of r\u003csub\u003eg\u003c/sub\u003e estimates adjusted for FDR. B. The outer cycle of the figure displays the heritability estimates for each of the 27 traits. The inner cycle presents the \u003cem\u003ep-\u003c/em\u003evalues of h\u003csub\u003e2\u003c/sub\u003e estimates adjusted for FDR.\u003c/p\u003e","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/9dbcd718f4bbcc0acea4486d.png"},{"id":61860667,"identity":"3c93a872-7fdc-4408-90b5-b221ff231303","added_by":"auto","created_at":"2024-08-06 10:47:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":595900,"visible":true,"origin":"","legend":"\u003cp\u003eTwo-sample Mendelian analysis results. A. Suggestive causal effects of 6 traits on NAFLD. B. The putative causal relationships between 6 traits and NAFLD using GSMR. The x-axis displays the effect sizes (bzx) of independent lead SNPs from the exposure's GWAS, while the y-axis shows the SNP GWAS effect sizes (bzy) for the outcome. Each point represents an independent lead SNP, with error bars indicating 95% confidence intervals for the SNP effect size. A dotted line, with a slope representing the estimated effect of the exposure on the outcome (bxy) and an intercept of zero, illustrates the potential causal direction between each trait and NAFLD.\u003c/p\u003e","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/cab75fdbc05ab3fa12d4906e.png"},{"id":61860665,"identity":"07af19e3-ce14-4a3c-8a69-08ee5beb389d","added_by":"auto","created_at":"2024-08-06 10:47:56","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":55315,"visible":true,"origin":"","legend":"\u003cp\u003eThe results of mediation analysis. The β value and P value calculated using the Inverse Variance Weighted(IVW) method.\u003c/p\u003e","description":"","filename":"OnlineFigure6.png","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/e4f7f526baf545994157e745.png"},{"id":62106293,"identity":"ccd19472-e67c-4c38-9a77-c0ae6c012d5c","added_by":"auto","created_at":"2024-08-09 10:52:04","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1395584,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/5ba42aad-195a-421c-b6ba-418f6945ebda.pdf"},{"id":61861150,"identity":"d2d2d1ca-c759-4db9-b67d-e932c5e67205","added_by":"auto","created_at":"2024-08-06 10:55:56","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":79808,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4722888/v1/5ac40a6730c47841f89bf57e.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"NAFLD’s Predisposion: insight from phenome-wide association and Mendelian Randomization","fulltext":[{"header":"Introduction","content":"\u003cp\u003eNon-alcoholic fatty liver disease (NAFLD) is a growing health problem worldwide, characterized by the accumulation of fat in liver cells even in individuals who consume little or no alcohol \u003csup\u003e1\u003c/sup\u003e. It is a multifactorial disease in which genetic, environmental and lifestyle factors combine to contribute to its pathogenesis and progression \u003csup\u003e2\u003c/sup\u003e. An in-depth understanding of the complex relationships between NAFLD and its associated phenotypes is essential for the development of targeted interventions and treatments. Recent advances in the fields of genomics and epidemiology have provided new tools to analyze these relationships. An GWAS meta-analysis found 17 genetic risk alleles shared at 17 statistically independent loci that were associated with NAFLD \u003csup\u003e3\u003c/sup\u003e. However, a limited number of functional genomics studies reveal the biological consequences associated with single nucleotide polymorphisms (SNPs) through in vitro and in vivo analyses. In addition to traditional GWAS approaches, the use of polygenic score, Mendelian randomization (MR), gene pathway analyses, and PheWAS can expand our understanding of NAFLD pathogenesis.\u003c/p\u003e\n\u003cp\u003ePhenome-Wide Association Studies (PheWAS) are strategically devised to explore across a multitude of phenotypes for identifying associations with genetic variants \u003csup\u003e4\u003c/sup\u003e. This methodology illuminates the potential influences of these variants on an extensive array of diseases and traits. The implementation of PheWAS has been significantly advanced by the advent of large-scale biobanks and genomic consortia, which facilitate comprehensive screenings across a wide range of clinical outcomes. By expanding the range of phenotypic expressions linked to a particular genetic alteration, performing Phenome-Wide Association Studies (PheWAS) on a Single Nucleotide Polymorphism (SNP) known to be related to a certain illness can uncover possible concurrent disorders. It can also identify characteristics that could play a role in the linkage between the SNP and the illness in question. For instance, a data-driven PheWAS study conducted within the UK Biobank explored the correlation between elevated Body Mass Index (BMI) and various disease outcomes \u003csup\u003e5\u003c/sup\u003e. This study utilized genetic risk scores and assessed causal relationships through MR analysis. It highlighted the multifaceted impact of BMI on numerous phenotypes, underscoring the utility of PheWAS in understanding complex traits. However, current research has not yet leveraged PheWAS to identify other \"intermediate traits\" that connect NAFLD risk alleles with the causal mechanisms underlying NAFLD pathogenesis. Therefore, the integration of PheWAS and MR presents a promising approach in NAFLD research, offering a deeper understanding of the genetic and phenotypic framework linked to NAFLD. This approach provides valuable information regarding its etiology, progression, and potential therapeutic targets. Our study aims to systematically explore the genetic basis of NAFLD and its wide-ranging phenotypic consequences using these methodologies.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThe datasets used were summary data; all informed consent and ethical approval were obtained in the original studies.\u003c/p\u003e\n\u003ch3\u003eSelection of NAFLD risk alleles\u003c/h3\u003e\n\u003cp\u003eWe compiled a selection of genetic variants frequently connected with NAFLD, ensuring they had a minor allele frequency of at least 1% among populations of European descent, to be examined in our PheWAS study. The selection drew upon the latest and most extensive genome-wide association study (GWAS) meta-analysis focused on NAFLD at that point, which integrated data from four cohorts with electronic health records confirming NAFLD in individuals of European background (comprising 3,584 cases and 621,081 controls) \u003csup\u003e3\u003c/sup\u003e. For the PheWAS inquiries, we progressed with 17 independent variants linked to NAFLD. This set encompassed 11 SNPs initially pinpointed through GWAS of NAFLD, along with an additional 6 new variants whose associations were validated in that same report.\u003c/p\u003e\n\u003ch3\u003eControl SNP set\u003c/h3\u003e\n\u003cp\u003ePrior studies have indicated that beyond a ratio of one case for every four controls, the gain in statistical power tends to be negligible \u003csup\u003e6\u003c/sup\u003e. Threrefore, for the 17 NAFLD risk SNPs, we generated a set of control SNPs at a 1:4 ratio using SNPsnap from the 1000 Genomes Project. We meticulously chose four control SNPs to correspond with those associated with NAFLD, adhering to multiple matching parameters: minor allele frequency within a 5% margin, gene density in the vicinity within a 50% range, proximity to the closest gene also within a 50% range, and to approximate the extent of the haplotype block, the count of surrounding SNPs in linkage disequilibrium (LD) with an R2 of 0.50 or above, allowing for a 50% variance.\u003c/p\u003e\n\u003ch3\u003ePheWAS analyses\u003c/h3\u003e\n\u003cp\u003eGeneATLAS database compiles GWAS data for 118 quantitative and 660 binary traits. The database catalogs associations with 9,113,133 genetic variants—both genotyped and imputed—from 452,264 UK Biobank participants of European ancestry. It supports queries for genetic or phenotypic data to assess genotype-phenotype associations.\u003c/p\u003e\n\u003cp\u003eIn this study, we used the GeneATLAS to identify associations between traits and established NAFLD risk SNPs, along with a control set of 68 SNPs identified through SNPsnap. For each queried variant, we extracted summary statistics related to associated traits from GeneATLAS for subsequent analysis. Following the methodology outlined by Semmes et al., SNP-trait linkages that showed a nominal level of significance (\u003cem\u003ep\u003c/em\u003e-value less than 0.01) were advanced for further trait-enrichment analysis \u003csup\u003e7\u003c/sup\u003e. This process was designed to compare traits associated with NAFLD-linked SNPs against those linked with the control SNP set.\u003c/p\u003e\n\u003cp\u003eGiven the exploratory nature of PheWAS, the significance threshold for SNP-trait associations was determined with a degree of arbitrariness, provided it remained consistent across all analyzed SNP sets. This approach allowed for a uniform standard for comparison and analysis, yet with the flexibility to accommodate the intrinsic variability and genetic correlations among multiple traits studied. For instance, considering the high genetic correlation between traits such as weight and BMI, or reticulocyte percentage and reticulocyte count, employing a more stringent \u003cem\u003ep-\u003c/em\u003evalue threshold could potentially be overly conservative. Such a strict threshold might overlook SNP-trait associations that, despite not meeting the highest levels of statistical stringency, could still hold biological relevance and significance in the context of NAFLD. This rationale underscores the importance of balancing statistical rigor with the practical considerations of genetic correlation and trait complexity in PheWAS analyses.\u003c/p\u003e\n\u003cp\u003ePheWAS analyses were conducted to explore the associations between 17 NAFLD-specific SNPs and a broad spectrum of 778 phenotypic traits. These associations were then benchmarked against the outcomes of a PheWAS analysis for a control set of SNPs, utilizing R version 4.1.3 for all statistical computations. To ascertain whether specific traits were enriched in association with the NAFLD risk variants compared to the control SNP set, we employed Fisher's exact test. This approach enabled us to compare how often specific traits occurred in association with SNPs related to NAFLD as well as with the control SNPs, aiming to identify significant trait enrichments linked to the known NAFLD risk variants.\u003c/p\u003e\n\u003cp\u003eTo account for the challenge of multiple comparisons inherent in such a broad analysis, the \u003cem\u003ep-\u003c/em\u003evalues obtained from Fisher's exact test for trait enrichment were subjected to Bonferroni correction. This adjustment ensures that the significance of our findings reflects a balanced consideration of both the potential for genuine associations and the statistical risk of type I errors due to the extensive number of tests performed.\u003c/p\u003e\n\u003ch3\u003eLinkage disequilibrium score (LDSC) analysis\u003c/h3\u003e\n\u003cp\u003eThe LDSC method serves as a powerful analytical tool for unraveling the genetic underpinnings of complex human traits. Utilizing summary-level data from GWAS, LDSC enables the quantification of heritability attributable to common genetic variants for diseases and facilitates the assessment of genetic overlap between traits. Our research concentrated on calculating the genetic correlation (r\u003csub\u003eg\u003c/sub\u003e) among NAFLD and related phenotypic traits, shedding light on the mutual genetic foundation that could be influencing these conditions. Further, we assessed the heritability (h\u003csup\u003e2\u003c/sup\u003e) of these phenotypes, which reflects the proportion of phenotypic variance that can be attributed to genetic factors. This analysis extended to computing the genetic correlations among the phenotypes themselves, aiming to identify potential genetic bridges that connect various traits within the context of NAFLD. To ensure the robustness of our findings, we applied the False Discovery Rate (FDR) correction to the \u003cem\u003ep-\u003c/em\u003evalues obtained for both the r\u003csub\u003eg\u003c/sub\u003e and h\u003csup\u003e2\u003c/sup\u003e estimates. The developers of the method have made these scores accessible through GitHub (https://github.com/bulik/ldsc).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTwo-sample mendelian randomization (TSMR) analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBidirectional TSMR analyses were performed to investigate the potential causative links between NAFLD and the phenotypes pinpointed via PheWAS, adhering to the STROBE-MR recommendations. The GWAS summary statistics for NAFLD were sourced from an electronic health record-based GWAS meta-analysis conducted by Nooshin Ghodsian and colleagues \u003csup\u003e8\u003c/sup\u003e, with data on other associated phenotypes detailed in Supplementary Table 1. In our study, we employed distinct GWAS meta-analysis data sources for the selection of NAFLD risk alleles and the TSMR analysis. This strategic approach was driven by our objectives for each analysis phase. For the Phewas analysis, our goal was to incorporate as many risk alleles as possible to comprehensively identify potential associations with NAFLD, leveraging a wider array of genetic variants. Conversely, for the TSMR process, the emphasis was on maximizing the total sample size and the number of SNPs to enhance the statistical power and robustness of our causal inference analysis.\u003c/p\u003e\n\u003cp\u003eIn this study, we primarily utilized the Generalized Summary-data-based Mendelian Randomization (GSMR) method in Genome-wide Complex Trait Analysis\u003c/p\u003e\n\u003cp\u003e(GCTA) software v1.94.1. This choice was motivated by GSMR's capability not only to estimate the causal effects of exposure factors on outcomes but also to identify and mitigate the effects of pleiotropy \u003csup\u003e9\u003c/sup\u003e. We chose near-independent SNPs from GWAS data using a linkage disequilibrium (LD) threshold of r² = 0.05 within a 1 Mb region and a significance threshold of p \u0026lt; 5.0 × 10^−8. The selection was based on LD estimates from the 1000G project's European population dataset, employing PLINK v2.0.0 alpha. In instances where a phenotype had fewer than 10 independent lead SNPs meeting the rigorous GWAS significance level, we lowered the threshold to p \u0026lt; 1.0 × 10^−5 to maintain an adequate pool of SNPs for our analysis. To mitigate the impact of horizontal pleiotropy on our results, we used heterogeneity in dependent instrument (HEIDI)-outlier detection as part of the GSMR \u0026nbsp;method. This step, taken before the analysis, involved applying the default HEIDI-outlier significance threshold of 0.01 to filter out unsuitable SNPs. Additionally, to assess the robustness of our results, we employed several other MR methods, including Inverse Variance Weighted (IVW), Weighted Median, and Weighted mode estimation with the R package “TwoSampleMR”. Furthermore, we performed MR-Egger regression to identify instances of horizontal pleiotropy, and MR-PRESSO was used to adjust for bias introduced by pleiotropic genetic variants. In situations where the GWAS data sources did not meet the computational requirements for GSMR in reverse MR analyses, alternative methods were utilized.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMediation analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our research, we utilized a two-sample Mendelian randomization strategy to pinpoint phenotypes with a causal link to NAFLD. The aim was to delve deeper into the mediators through which these phenotypes precipitate the onset of NAFLD. Central to this investigation are inflammatory mediators, which act as a critical bridge connecting various metabolic and immune pathways associated with NAFLD. These mediators orchestrate a series of mechanisms, including insulin resistance \u003csup\u003e10\u003c/sup\u003e, oxidative stress \u003csup\u003e11\u003c/sup\u003e, and the activation of the innate immune system \u003csup\u003e12\u003c/sup\u003e. Their actions contribute significantly to the progression of liver damage, evolving from simple steatosis to more severe conditions like Non-Alcoholic Steatohepatitis (NASH), fibrosis, and, in some cases, cirrhosis \u003csup\u003e13\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eUtilizing data from a study that explored genome variant associations with 91 plasma proteins across 14,824 European participants, it is possible to uncover potential intermediary variables that mediate various phenotypes leading to NAFLD\u0026nbsp;\u003csup\u003e14\u003c/sup\u003e. Our initial step involved using TSMR analysis to identify factors causally related to NAFLD, focusing on inflammatory proteins as exposure factors. Next, we treated the phenotypes causally linked to NAFLD as exposure factors and the inflammatory proteins, which were established as causally related to NAFLD, as outcome variables. In cases where exposures had a significant impact on outcomes, the 'coefficient product' approach was implemented to evaluate the indirect effects exerted by these phenotypes on NAFLD via mediating pathways. This method involves calculating the mediator's effect by multiplying the influence of the exposure on the mediator with the mediator's impact on the outcome. To understand the contribution of each mediator to the overall outcome, we calculated the proportion of the mediating effect relative to the total effect of the exposure on the outcome using the formula (β1 × β2/β3).\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e: Research Framework. This figure delineates our methodology for exploring associations between phenotypes and established disease risk variants, aiming to pinpoint phenotypes for further LDSC and MR analysis. When applied to NAFLD, this strategy led to the identification of 31 phenotypes significantly associated with NAFLD. It also facilitated subsequent median analyses to assess how variants linked to inflammatory proteins contribute to the risk of NAFLD.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSNPs associated with NAFLD, validated in previous GWAS meta-analyses, and incorporated into our enrichment analysis within the PheWAS framework.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePosition (Chr:BP)\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ersID\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMAF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFunction \u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNearest gene\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22:44324730\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers738408\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePNPLA3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19:19379549\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers58542926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTM6SF2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19:45411941\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers429358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003emissense\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAPOE\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2:27730940\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers1260326\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eGCKR\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8:126506694\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers112875651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTRIB1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10:113947040\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers4918722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eGPAM\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1:220970028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers2642438\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMARC1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4:100505326\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers7661964\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u0026rsquo; UTR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMTTP\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9:132566666\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers7029757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTOR1B\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4:100239319\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers1229984\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eADH1B\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e16:53813367\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers17817449\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFTO\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2:165555539\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers79953491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCOBLL1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19:7218635\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers112630404\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eINSR\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19:54677001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers626283\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintergenic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTMC4/MBOAT7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e17:17979099\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers4561528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSREBF1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9:10462423\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers10756038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eintron variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePTPRD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11:823586\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ers140201358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eexonic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePNPLA2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003e\u003csup\u003ea\u003c/sup\u003e Position in GRCh37/hg19.\u003c/p\u003e \u003cp\u003e\u003csup\u003eb\u003c/sup\u003e Minor allele frequency in European-ancestry individuals from the 1000 genomes project.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003ePhewas analysis\u003c/h2\u003e \u003cp\u003ePheWAS were conducted utilizing the UK Biobank GeneATLAS database to assess the association of each NAFLD-linked variant and control variants, with a comprehensive array of 778 traits. These traits could be organized into 24 distinct categories.\u003c/p\u003e \u003cp\u003eAmong the 68 control SNPs evaluated, six variants (rs186719489, rs111571790, rs139635278, rs962745, rs118065116, rs61952239) were not present in the GeneATLAS database. Consequently, our analysis proceeded with 17 test SNPs and 62 control SNPs included in the enrichment analysis.\u003c/p\u003e \u003cp\u003eWithin the GeneATLAS database, we found that 329/778 traits exhibited a nominal association (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) with at least one of the 17 SNPs that have been linked to NAFLD. These traits were further subjected to enrichment analysis against the control SNP set, as detailed in Supplementary Table S2.\u003c/p\u003e \u003cp\u003eOut of the traits assessed, 128 were determined to have a higher probability of association with the SNPs linked to NAFLD than with the control SNPs (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Following Bonferroni correction, 31 of these traits retained their significance, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Notably, the category with the highest number of significantly enriched traits was physical and body measurements, encompassing 10 distinct phenotypes, followed by genetic and biochemical markers.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eLDSC analysis\u003c/h2\u003e \u003cp\u003eFrom the 31 phenotypes significantly associated with NAFLD, we excluded E10-E14 Diabetes mellitus due to unclear definitions. Instead, we incorporated Type 1 and Type 2 diabetes for further analysis. K76 Other diseases of liver, due to overlapping diagnostic codes with NAFLD, E78 Disorders of lipoprotein metabolism and other lipidaemias, I20-I25 Ischaemic heart diseases and E11 Non-insulin-dependent diabetes mellitus, all of which had no related GWAS data available, were also excluded. Consequently, 27 traits were included in the subsequent analysis.\u003c/p\u003e \u003cp\u003eInitially, we conducted pairwise LDSC analysis on the selected 27 traits, identifying several pairs with significant genetic correlations, as depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. Notably, there were pronounced genetic correlations between certain physical and body measurements and metabolic diseases. Specifically, Type 2 diabetes, Arm fat-free mass, and BMI exhibited substantial positive r\u003csub\u003eg\u003c/sub\u003e. Conversely, Impedance of the whole body demonstrated negative genetic correlations with the aforementioned indicators.\u003c/p\u003e \u003cp\u003eSubsequently, we analyzed the heritability of these 27 traits as well as their genetic correlation with NAFLD. As depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, the LDSC analysis revealed a strong positive correlation of NAFLD with several key traits: BMI showing a genetic correlation of 0.73, Trunk fat mass with r\u003csub\u003eg\u003c/sub\u003e = 0.67, Type 2 diabetes with r\u003csub\u003eg\u003c/sub\u003e = 0.86, and weight with r\u003csub\u003eg\u003c/sub\u003e = 0.57. Additionally, a negative correlation of NAFLD with Impedance of the whole body (r\u003csub\u003eg\u003c/sub\u003e = -0.31) and Neutrophil count (r\u003csub\u003eg\u003c/sub\u003e = -0.28) was also observed. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB highlights the traits with the highest heritability estimates, including Whole body fat-free mass (h\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.31), Trunk fat-free mass (h\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.31), and Trunk predicted mass (h\u003csub\u003e2\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.31).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eTSMR analysis\u003c/h2\u003e \u003cp\u003eWe further investigated bidirectional causal associations between 27 traits and NAFLD. Our ability to explore causal associations using GSMR, with NAFLD as the exposure and other traits as outcomes, was limited due to an insufficient number of lead SNPs for NAFLD (n\u0026thinsp;\u0026lt;\u0026thinsp;10 at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;1.0 \u0026times; 10^\u0026minus;5). Consequently, in the reverse MR analysis, we employed five alternative methods, with the IVW method serving as the primary analysis.\u003c/p\u003e \u003cp\u003eAs illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA-B and Supplementary table 4, we identified unidirectional progressive effects of BMI, Trunk fat mass, Weight, and Type 2 diabetes on NAFLD. Conversely, Impedance of the whole body was identified as a protective factor against NAFLD. The GSMR results for each of these phenotypes were corroborated by at least two other methods.\u003c/p\u003e \u003cp\u003eFurthermore, we observed strong bidirectional progressive effects of Alcohol intake frequency on NAFLD, indicating a reciprocal influence between alcohol consumption and NAFLD risk. This bidirectional MR analysis was validated in at least two methods besides the primary analysis, reinforcing the complex interplay between alcohol intake and NAFLD (Supplementary Table\u0026nbsp;5).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eMedian analysis\u003c/h2\u003e \u003cp\u003eUtilizing the 'product of coefficients' approach, we pinpointed 2 potential causal routes that could influence the development of NAFLD (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e and Table S6). The effect of BMI on NAFLD was partially mediated by Fibroblast growth factor 21 (FGF21) (indirect effects\u0026thinsp;=\u0026thinsp;0.030, mediated proportion: 6.48%). Trunk fat mass might increase the risk of NAFLD by increasing Interleukin-10 receptor subunit beta (IL-10RB) (indirect effects\u0026thinsp;=\u0026thinsp;0.014, mediated proportion: 4.00%).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe adopted an innovative and corroborated model to sift through extant GWAS and PheWAS datasets, aiming to identify characteristics correlated with NAFLD susceptibility. Initially, we curated SNPs linked to NAFLD from extensive comprehensive GWAS meta-analysis and paired these risk variants to control SNPs. Through PheWAS conducted on these SNPs utilizing UK Biobank data, 31 traits were identified as significantly enriched for associations with NAFLD risk SNPs. LDSC further enabled us to determine the genetic correlations among these 31 traits, unveiling both positive and negative genetic correlations with NAFLD. TSMR analysis, with GSMR as the primary method, implicated six traits as playing a causal role in NAFLD pathogenesis.\u003c/p\u003e \u003cp\u003ePheWAS results indicated that under the category of physical and body measurements, ten traits were significantly enriched for associations with NAFLD. However, only BMI, Trunk fat mass, and weight were suggested by TSMR analyses to exhibit vertical pleiotropy, directly contributing to NAFLD risk. Elevated levels of these indicators are indicative of obesity, a condition prevalent in up to 80% of NAFLD patients. Obesity is considered as a state of chronic low-grade inflammation and is associated with various complications including NAFLD.\u003c/p\u003e \u003cp\u003eThe mechanisms through which obesity leads to NAFLD involve complex interactions among metabolic processes, including insulin resistance, altered lipid metabolism, and inflammation. Specifically, obesity-induced insulin resistance is a central factor that exacerbates hepatic fat accumulation through increased hepatic fatty acid influx and lipogenesis, coupled with decreased fatty acid oxidation \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Additionally, obesity leads to elevated levels of free fatty acids released by adipose tissue, which further augment hepatic uptake and synthesis of lipids, potentially reducing β-oxidation and contributing to hepatic fat accumulation \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Inflammatory factors, notably TNF-α, IL-1β and IL-6 produced especially by visceral fat, can promote subacute hepatic inflammation and fibrosis through various signaling pathways \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Moreover, obesity alters the secretion profile of hormones and cytokines by adipocytes, such as reduced adiponectin and increased leptin levels, impacting overall metabolic state and liver metabolism \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eOur study highlights the nuanced relationship between different physical measurement indicators related to body composition and NAFLD. While BMI, Trunk fat mass, and Weight were identified as causal factors for NAFLD, other body physical measurements such as Left arm fat-free/predicted mass, Right arm fat-free/predicted mass, and Hip circumference, despite their association with NAFLD in PheWAS results and positive genetic correlations as shown by LDSC, did not emerge as causal in TSMR analyses. This discrepancy underscores the complexity of NAFLD's etiology, suggesting that not all obesity-related markers carry the same risk for NAFLD. Specifically, our findings align with the notion that central obesity, as reflected by Trunk fat mass, plays a more pivotal role in NAFLD pathogenesis compared to other fat distribution patterns. For instance, studies have shown that visceral fat, rather than subcutaneous fat, is more metabolically active and contributes significantly to hepatic steatosis and insulin resistance, thereby increasing the risk of NAFLD \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Furthermore, the differential impact of fat depots on NAFLD risk underscores the importance of targeted interventions that reduce visceral fat to mitigate NAFLD risk, beyond the general reduction of body weight or BMI. Our study, therefore, emphasizes the significance of considering specific fat distribution patterns in NAFLD risk assessment and management strategies.\u003c/p\u003e \u003cp\u003eWhole-body impedance is a valuable tool for assessing body composition, including distinctions between fat and lean tissues. It is generally observed that higher whole-body impedance values correlate with a lower amount of body fat \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. This relationship underpins our finding that whole-body impedance acts as a protective factor against NAFLD.\u003c/p\u003e \u003cp\u003eIn the context of clinical outcomes, the Fat-Free Mass Index (FFMI), derived from whole-body impedance measurements, shows promising utility. It has demonstrated comparable, and in some cases superior, predictive power for NAFLD risk compared to traditional metrics such as BMI \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. This suggests that incorporating whole-body impedance assessments into clinical practice could provide a more nuanced understanding of NAFLD risk, beyond conventional obesity measurements .\u003c/p\u003e \u003cp\u003eNotably, the association between bioelectrical impedance-derived metrics and liver health has been observed in various studies. For instance, research has shown that bioelectrical impedance analysis can effectively reflect changes in body composition, including muscle mass, which is inversely associated with the severity of hepatic steatosis in patients with NAFLD \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. This underscores the potential of whole-body impedance measurements not only as a diagnostic tool but also as a predictor of NAFLD progression, highlighting the importance of lean mass preservation in NAFLD management strategies.\u003c/p\u003e \u003cp\u003eOur findings highlight a bidirectional causal relationship between alcohol intake frequency and NAFLD. The effects of Alcohol intake frequency on NAFLD (OR\u0026thinsp;=\u0026thinsp;1.42) were stronger than those of NAFLD on Alcohol intake frequency (OR\u0026thinsp;=\u0026thinsp;1.02). The odds ratio of 1.42 suggests that with each ascending level of alcohol intake frequency\u0026mdash;from 'Never,' 'Special occasions only,' '1\u0026ndash;3 times a month,' 'Once or twice a week,' 'Three or four times a week,' to 'Daily or almost daily'\u0026mdash;the risk of developing NAFLD increases by 42%. Conversely, the presence of NAFLD is associated with a slight increase in the frequency of alcohol consumption. However, this increase is relatively modest when compared to the impact of alcohol intake frequency on NAFLD risk. Traditionally, heavy drinking has been unequivocally linked to liver damage, while the effects of non-heavy consumption remain contentious \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. A notable prospective cohort study suggested that even moderate drinking might exacerbate fibrosis in NAFLD patients, underscoring potential risks of any alcohol intake in this population \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Contrarily, some cross-sectional studies report no adverse effects from moderate alcohol use on NAFLD, though these findings could be skewed by factors like temporal uncertainty and reverse causality, where sicker patients might abstain from drinking \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFGF21, a hormone primarily expressed in the liver, plays a pivotal role in regulating lipid and glucose metabolism, boasting capabilities to enhance insulin sensitivity and promote fatty acid oxidation. Moreover, FGF21 exhibits potential in inhibiting the development of NAFLD through its promotion of fatty acid oxidation and its ability to reduce hepatic fat accumulation via insulin-independent pathways \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Despite the therapeutic potential of FGF21 in treating metabolic diseases, elevated levels of FGF21 in individuals with obesity and NAFLD suggest a possible dysfunction or resistance to FGF21 signaling in these conditions \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Our study found that FGF21 acts as a mediating factor in the pathway from BMI to NAFLD, further emphasizing the significance of FGF21 in metabolic health. This indicates that an upregulation of FGF21 expression, in response to increased body weight, might be the body's attempt to cope with metabolic stress induced by obesity through enhancing lipid oxidation and improving metabolic health. However, the rise in FGF21 levels may also reflect a failure of metabolic adaptation.\u003c/p\u003e \u003cp\u003eAdditionally, through mediation MR analysis, IL-10RB was demonstrated to mediate the pathway through which trunk fat mass leads to NAFLD. IL-10RB, an integral part of the IL-10 signaling cascade, is instrumental in regulating immune responses and suppressing inflammation in adipose tissue and insulin resistance \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Recent studies have shown that IL-10 can directly inhibit the thermogenesis of adipocytes through a STAT3-dependent signaling pathway, and germline deletion of IL-10 can protect mice from insulin resistance and diet-induced obesity (DIO) \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Furthermore, the specific deletion of IL-10 or Blimp-1 in Treg cells can improve insulin sensitivity and DIO, underscoring that the inhibitory effect of Treg cells through IL-10 secretion extends beyond interactions among immune cells to also include the suppression of the beiging process in non-immune cells like adipocytes. Based on this, we infer that in the context of obesity and NAFLD, excessive fat accumulation, especially in the trunk, is closely associated with an increased inflammatory state. The overexpression of IL-10 under chronic inflammatory levels affects the functionality of adipocytes, disrupting the inflammatory balance and metabolic stability of adipose tissue, thereby promoting the development of NAFLD.\u003c/p\u003e \u003cp\u003eOur research has certain constraints and justifiable issues regarding the combined GWAS-PheWAS methodology we have adopted. Specifically, our enrichment analysis of NAFLD risk SNPs across 778 traits involved Bonferroni correction to adjust for multiple comparisons. While this rigorous correction method minimizes the risk of false positives, it may also reduce the sensitivity to detect traits potentially sharing genetic variations with NAFLD, potentially excluding phenotypes of relevance.\u003c/p\u003e \u003cp\u003eAdditionally, our study's analyses were exclusively based on traits available in the U.K. Biobank, which predominantly features a European population. Consequently, the findings may not universally apply across different ethnic groups. For instance, while our results identify obesity-related metrics such as BMI as risk factors for NAFLD, it is important to note that not all obese individuals develop NAFLD. More critically, NAFLD can also occur in individuals who are not obese. Although NAFLD in non-obese individuals has been observed across various ethnicities, including children and adults, it is reported more frequently in Asian populations, even when using strict, ethnicity-specific BMI criteria for obesity \u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. This suggests that conducting similar analyses in diverse ethnic backgrounds could yield varying insights, underscoring the importance of considering ethnic diversity in understanding NAFLD's genetic predispositions.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn summary, our integration of GWAS and PheWAS datasets has illuminated the multifaceted etiology of NAFLD, uncovering both horizontal and vertical pleiotropy across several traits. Importantly, our findings implicate FGF21 and IL-10RB as significant players in the pathogenesis of NAFLD. These insights pave the way for more in-depth mechanistic studies aimed at understanding the specific contributions of these traits to NAFLD pathogenesis and exploring their interconnected relationships. The knowledge gained holds promise for informing the development of more targeted prevention and treatment strategies for NAFLD.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis research and the associated article processing charges were funded by the Sichuan Provincial Science and Technology Plan Project of Science and Technology Bureau of Sichuan, grant number 2023JDRC0092.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eH.Y.J. and Q.D. contribute equally to this work. Conceptualization, H.Y.J. and H.Y.Y; methodology, Q.D.; software, H.Y.J.; formal analysis, H.Y.J. and Q.D.; writing\u0026mdash;original draft preparation, H.Y.J. and Q.D.; writing\u0026mdash;review and editing, Q.D.; visualization, H.Y.J.; supervision, T.S. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData is provided within the manuscript or supplementary information files\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eYounossi, Z. M. \u003cem\u003eet al.\u003c/em\u003e Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. \u003cem\u003eHepatology\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, 73\u0026ndash;84 (2016).\u003c/li\u003e\n \u003cli\u003eEstes, C., Razavi, H., Loomba, R., Younossi, Z. \u0026amp; Sanyal, A. J. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. \u003cem\u003eHepatology\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 123\u0026ndash;133 (2018).\u003c/li\u003e\n \u003cli\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Genome-wide association meta-analysis identifies 17 loci associated with nonalcoholic fatty liver disease. \u003cem\u003eNat Genet\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 1640\u0026ndash;1650 (2023).\u003c/li\u003e\n \u003cli\u003eBastarache, L., Denny, J. C. \u0026amp; Roden, D. M. Phenome-Wide Association Studies. \u003cem\u003eJAMA\u003c/em\u003e \u003cstrong\u003e327\u003c/strong\u003e, 75\u0026ndash;76 (2022).\u003c/li\u003e\n \u003cli\u003eHypp\u0026ouml;nen, E., Mulugeta, A., Zhou, A. \u0026amp; Santhanakrishnan, V. K. A data-driven approach for studying the role of body mass in multiple diseases: a phenome-wide registry-based case-control study in the UK Biobank. \u003cem\u003eLancet Digit Health\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, e116\u0026ndash;e126 (2019).\u003c/li\u003e\n \u003cli\u003eHennessy, S., Bilker, W. B., Berlin, J. A. \u0026amp; Strom, B. L. Factors influencing the optimal control-to-case ratio in matched case-control studies. \u003cem\u003eAm J Epidemiol\u003c/em\u003e \u003cstrong\u003e149\u003c/strong\u003e, 195\u0026ndash;197 (1999).\u003c/li\u003e\n \u003cli\u003eSemmes, E. C. \u003cem\u003eet al.\u003c/em\u003e Leveraging Genome and Phenome-Wide Association Studies to Investigate Genetic Risk of Acute Lymphoblastic Leukemia. \u003cem\u003eCancer Epidemiol Biomarkers Prev\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1606\u0026ndash;1614 (2020).\u003c/li\u003e\n \u003cli\u003eGhodsian, N. \u003cem\u003eet al.\u003c/em\u003e Electronic health record-based genome-wide meta-analysis provides insights on the genetic architecture of non-alcoholic fatty liver disease. \u003cem\u003eCell Rep Med\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 100437 (2021).\u003c/li\u003e\n \u003cli\u003eZhu, Z. \u003cem\u003eet al.\u003c/em\u003e Causal associations between risk factors and common diseases inferred from GWAS summary data. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 224 (2018).\u003c/li\u003e\n \u003cli\u003eAsrih, M. \u0026amp; Jornayvaz, F. R. Inflammation as a potential link between nonalcoholic fatty liver disease and insulin resistance. \u003cem\u003eJ Endocrinol\u003c/em\u003e \u003cstrong\u003e218\u003c/strong\u003e, R25-36 (2013).\u003c/li\u003e\n \u003cli\u003eXu, D. \u003cem\u003eet al.\u003c/em\u003e The Role of Nrf2 in Liver Disease: Novel Molecular Mechanisms and Therapeutic Approaches. \u003cem\u003eFront Pharmacol\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1428 (2018).\u003c/li\u003e\n \u003cli\u003eArrese, M., Cabrera, D., Kalergis, A. M. \u0026amp; Feldstein, A. E. Innate Immunity and Inflammation in NAFLD/NASH. \u003cem\u003eDig Dis Sci\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 1294\u0026ndash;1303 (2016).\u003c/li\u003e\n \u003cli\u003eBuzzetti, E., Pinzani, M. \u0026amp; Tsochatzis, E. A. The multiple-hit pathogenesis of non-alcoholic fatty liver disease (NAFLD). \u003cem\u003eMetabolism\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e, 1038\u0026ndash;1048 (2016).\u003c/li\u003e\n \u003cli\u003eZhao, J. H. \u003cem\u003eet al.\u003c/em\u003e Genetics of circulating inflammatory proteins identifies drivers of immune-mediated disease risk and therapeutic targets. \u003cem\u003eNat Immunol\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 1540\u0026ndash;1551 (2023).\u003c/li\u003e\n \u003cli\u003eMilić, S., Lulić, D. \u0026amp; \u0026Scaron;timac, D. Non-alcoholic fatty liver disease and obesity: biochemical, metabolic and clinical presentations. \u003cem\u003eWorld J Gastroenterol\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 9330\u0026ndash;9337 (2014).\u003c/li\u003e\n \u003cli\u003ePolyzos, S. A., Kountouras, J. \u0026amp; Zavos, C. Nonalcoholic fatty liver disease: the pathogenetic roles of insulin resistance and adipocytokines. \u003cem\u003eCurr Mol Med\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 299\u0026ndash;314 (2009).\u003c/li\u003e\n \u003cli\u003eBoden, G. \u003cem\u003eet al.\u003c/em\u003e Free fatty acids produce insulin resistance and activate the proinflammatory nuclear factor-kappaB pathway in rat liver. \u003cem\u003eDiabetes\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 3458\u0026ndash;3465 (2005).\u003c/li\u003e\n \u003cli\u003eMonsour, H. P., Frenette, C. T. \u0026amp; Wyne, K. Fatty liver: a link to cardiovascular disease--its natural history, pathogenesis, and treatment. \u003cem\u003eMethodist Debakey Cardiovasc J\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 21\u0026ndash;25 (2012).\u003c/li\u003e\n \u003cli\u003eFracanzani, A. L. \u003cem\u003eet al.\u003c/em\u003e Liver and Cardiovascular Damage in Patients With Lean Nonalcoholic Fatty Liver Disease, and Association With Visceral Obesity. \u003cem\u003eClin Gastroenterol Hepatol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 1604-1611.e1 (2017).\u003c/li\u003e\n \u003cli\u003eFoster, K. R. \u0026amp; Lukaski, H. C. Whole-body impedance--what does it measure? \u003cem\u003eAm J Clin Nutr\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, 388S-396S (1996).\u003c/li\u003e\n \u003cli\u003eElia, M. Body composition by whole-body bioelectrical impedance and prediction of clinically relevant outcomes: overvalued or underused? \u003cem\u003eEur J Clin Nutr\u003c/em\u003e \u003cstrong\u003e67 Suppl 1\u003c/strong\u003e, S60-70 (2013).\u003c/li\u003e\n \u003cli\u003eYodoshi, T. \u003cem\u003eet al.\u003c/em\u003e Impedance-based measures of muscle mass can be used to predict severity of hepatic steatosis in pediatric nonalcoholic fatty liver disease. \u003cem\u003eNutrition\u003c/em\u003e \u003cstrong\u003e91\u0026ndash;92\u003c/strong\u003e, 111447 (2021).\u003c/li\u003e\n \u003cli\u003eLiangpunsakul, S. \u0026amp; Chalasani, N. What should we recommend to our patients with NAFLD regarding alcohol use? \u003cem\u003eAm J Gastroenterol\u003c/em\u003e \u003cstrong\u003e107\u003c/strong\u003e, 976\u0026ndash;978 (2012).\u003c/li\u003e\n \u003cli\u003eChang, Y. \u003cem\u003eet al.\u003c/em\u003e Nonheavy Drinking and Worsening of Noninvasive Fibrosis Markers in Nonalcoholic Fatty Liver Disease: A Cohort Study. \u003cem\u003eHepatology\u003c/em\u003e \u003cstrong\u003e69\u003c/strong\u003e, 64\u0026ndash;75 (2019).\u003c/li\u003e\n \u003cli\u003eAjmera, V. H., Terrault, N. A. \u0026amp; Harrison, S. A. Is moderate alcohol use in nonalcoholic fatty liver disease good or bad? A critical review. \u003cem\u003eHepatology\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e, 2090\u0026ndash;2099 (2017).\u003c/li\u003e\n \u003cli\u003eLiu, J., Xu, Y., Hu, Y. \u0026amp; Wang, G. The role of fibroblast growth factor 21 in the pathogenesis of non-alcoholic fatty liver disease and implications for therapy. \u003cem\u003eMetabolism\u003c/em\u003e \u003cstrong\u003e64\u003c/strong\u003e, 380\u0026ndash;390 (2015).\u003c/li\u003e\n \u003cli\u003eInagaki, T. Research Perspectives on the Regulation and Physiological Functions of FGF21 and its Association with NAFLD. \u003cem\u003eFront Endocrinol (Lausanne)\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 147 (2015).\u003c/li\u003e\n \u003cli\u003eOuyang, W. \u0026amp; O\u0026rsquo;Garra, A. IL-10 Family Cytokines IL-10 and IL-22: from Basic Science to Clinical Translation. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 871\u0026ndash;891 (2019).\u003c/li\u003e\n \u003cli\u003eRajbhandari, P. \u003cem\u003eet al.\u003c/em\u003e IL-10 Signaling Remodels Adipose Chromatin Architecture to Limit Thermogenesis and Energy Expenditure. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e172\u003c/strong\u003e, 218-233.e17 (2018).\u003c/li\u003e\n \u003cli\u003eKim, D. \u0026amp; Kim, W. R. Nonobese Fatty Liver Disease. \u003cem\u003eClin Gastroenterol Hepatol\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 474\u0026ndash;485 (2017).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"nonalcoholic fatty liver disease, mendelian randomization, phenome-wide association study, linkage disequilibrium score regression","lastPublishedDoi":"10.21203/rs.3.rs-4722888/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4722888/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGenome-wide association studies (GWAS) meta-analysis have unveiled common single nucleotide polymorphisms (SNPs) associated with the increased risk of Non-alcoholic fatty liver disease (NAFLD). We conducted a phenome-wide association study (PheWAS) using data from the UK Biobank to further elucidate NAFLD-associated phenotypes and investigate the disease's underlying biology. A significant enrichment was found in 31 of 778 traits examined using 17 known NAFLD-risk SNPs along with 4:1 matched control SNPs. We explored genetic correlations and causal relationships by employing bidirectional Mendelian randomization (MR) and linkage disequilibrium score regression (LDSC). Notably, strong positive genetic correlations with NAFLD were observed for BMI (r\u003csub\u003eg\u003c/sub\u003e = 0.73), Trunk fat mass (r\u003csub\u003eg\u003c/sub\u003e = 0.67), Type 2 diabetes (r\u003csub\u003eg\u003c/sub\u003e = 0.86), and weight (r\u003csub\u003eg\u003c/sub\u003e = 0.57), while whole-body impedance (r\u003csub\u003eg\u003c/sub\u003e = -0.31) and neutrophil count (r\u003csub\u003eg\u003c/sub\u003e = -0.28) exhibited negative correlations. Our MR analysis demonstrated unidirectional effects of BMI (OR = 1.57), Trunk fat mass (OR = 1.40), Weight (OR = 1.22), whole-body Impedance (OR = 0.83), and Type 2 diabetes (OR = 1.42) on NAFLD risk. Intriguingly, bidirectional causal effects were identified between Alcohol intake frequency and NAFLD (OR\u003csub\u003eAlcohol intake frequency → NAFLD\u003c/sub\u003e = 1.42; OR\u003csub\u003eNAFLD → Alcohol intake frequency\u003c/sub\u003e = 1.02), suggesting a complex interplay. Furthermore, through intermediary MR analyses, we uncovered pathways mediated by FGF21 and IL-10RB, linking BMI and Trunk fat mass, respectively, to NAFLD development. These findings provide novel insights into the multifaceted genetic landscape of NAFLD, highlighting the importance of body composition, metabolic health, and lifestyle factors in its pathogenesis.\u003c/p\u003e","manuscriptTitle":"NAFLD’s Predisposion: insight from phenome-wide association and Mendelian Randomization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-08-06 10:47:51","doi":"10.21203/rs.3.rs-4722888/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1d24f3d4-1e5a-4b2d-90e6-4f6cd8af3669","owner":[],"postedDate":"August 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-08-13T06:10:08+00:00","versionOfRecord":[],"versionCreatedAt":"2024-08-06 10:47:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4722888","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4722888","identity":"rs-4722888","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}