The contribution of tissue-grouped BMI-associated gene sets to cardiometabolic-disease risk: a Mendelian randomization study

• Objective: This Mendelian randomization (MR) study aimed to determine if grouping BMI-associated genetic variants based on differential tissue expression profiles (from GTEx data) could identify distinct causal effects on Type 2 Diabetes Mellitus (T2DM) and Coronary Artery Disease (CAD).
• Key Finding: The study identified 17 tissue-grouped gene sets; however, all sets showed a similar association with increased risks of T2DM and CAD, with effect estimates comparable to those from randomly sampled genetic variants.
• Conclusion: The novel approach of clustering BMI genetic instruments by tissue expression did not provide additional insight into the role of specific tissues in the genetic risk for these cardiometabolic diseases.

PubMed: 32500151 DOI: 10.1093/ije/dyaa070 Overview generated by: Gemini 2.5 Flash, 28/11/2025

Key Findings

The study aimed to test a biology-based approach by grouping BMI-associated genetic variants according to the tissue in which they are differentially expressed, expecting to find evidence for distinct pathways leading to cardiometabolic disease risk.

• No Differential Risk based on Tissue Grouping: The central finding was that all identified tissue-grouped BMI-associated gene sets were similarly associated with increased risks of Type 2 Diabetes Mellitus (T2DM) and Coronary Artery Disease (CAD), despite being biologically distinct.
• Similarity to Random Samples: The distribution of effect estimates (Odds Ratios, ORs) for T2DM and CAD derived from the tissue-grouped sets was similar to the distribution obtained from randomly sampled sets of BMI-associated genetic variants.
• Lack of Additional Insight: The conclusion was that this novel, biology-based approach did not provide additional insight into the role of specific tissues in the genetic risk for specific cardiometabolic diseases (T2DM or CAD) due to overweight or obesity.
• Overall Risk Confirmed: Considering all BMI genetic instruments, an increase of 1 Standard Deviation (SD) in BMI was associated with an increased risk of T2DM (OR 2.71, 95% CI 2.49; 2.94) and CAD (OR 1.48, 95% CI 1.37; 1.59).

Study Design and Methods

Study Design

This was a two-sample Mendelian randomization (MR) study that used publicly available summary-level data from large-scale Genome-Wide Association Studies (GWAS) to assess causal relationships.

Selection and Grouping of Genetic Instruments

1. Selection of Exposure Variants: The exposure was Body Mass Index (BMI). Genetic variants were selected from the GWAS meta-analysis by Yengo et al. (2018), which initially identified 656 independent single nucleotide polymorphisms (SNPs) associated with BMI. After filtering and gene annotation, 634 mapped BMI-associated genes were used.
2. Tissue-Grouped Gene Set Identification: Overrepresentation tests were performed using Genotype-Tissue Expression (GTEx) version 8 data to identify genes that were differentially expressed (significantly up- or downregulated) in a given tissue compared with other tissues. This resulted in the identification of 17 partly overlapping tissue-grouped gene sets. Twelve of these sets were derived from brain areas (e.g., hypothalamus, amygdala), while others included the artery tibial, kidney cortex, spleen, and digestive system tissues.
3. MR Instrumental Variables: The individual SNPs corresponding to the genes within these 17 tissue-grouped sets were used as instrumental variables for the MR analyses. The genetic instrument mapped to TCF7L2 (rs7903146) was removed from all gene sets due to its known pleiotropic effect.

Outcome Data

• T2DM: Summary statistics from the DIAbetes Genetics Replication and Meta-analysis (DIAGRAM) consortium (74,124 cases and 824,006 controls).
• CAD: Summary statistics from the CARDIOGRAMplusC4D consortium (60,801 cases and 123,504 controls).
• Anthropometric Traits: Waist circumference and total body fat were also used as outcomes to explore underlying mechanisms.

Statistical Analysis

• Primary MR Method: Inverse-Variance-Weighted (IVW) analyses were used to combine the effects of individual genetic instruments. The results were reported as the change in Odds Ratio (OR) per 1 SD increase in BMI (which corresponds to 4.8 kg/m\(^2\)).
• Sensitivity Analyses: MR-Egger regression and Weighted Median-Estimator (WME) analyses were performed to test for and account for potential pleiotropy.

Conclusions and Recommendations

The investigation demonstrated that grouping BMI-associated genes based on tissue-specific expression (a biological mechanism) did not yield groups of variants with distinct causal effects on T2DM or CAD risk.

• Implication for Etiological Studies: The findings suggest that the genetic predisposition to cardiometabolic disease driven by general BMI is largely independent of the tissue-specific expression profiles examined, or that the current methodology for tissue-grouping is insufficient to detect fine-scale differences in downstream risk pathways.
• Limitation: A potential limitation noted by the authors is that the GTEx overrepresentation test does not account for the effect size of the genetic instruments on the exposure (BMI), meaning large and small effect variants are weighted equally in the clustering, which may have masked subtle differences. The substantial overlap between genes within the tissue-grouped gene sets was also cited as a limitation that may reduce specificity.