This review provides an overview of meta-analysis and Mendelian randomization (MR) in epidemiological research. MR uses genetic variants as instrumental variables to infer causal relationships between risk factors and health outcomes. Traditionally, MR combined results from separate studies using a small number of genetic variants, but recent advances allow the use of genome-wide association study (GWAS) summary data for large numbers of genetic variants. However, heterogeneity among causal estimates from different genetic variants may violate instrumental variable assumptions. The review explains the theoretical basis of MR, including instrumental variable theory, and describes methods for testing and adjusting for heterogeneity, such as random effects models, meta-regression, and robust regression. It also discusses the use of two-sample summary data MR, which allows the combination of GWAS summary data from different studies without requiring individual-level data. The review highlights the importance of addressing potential biases, such as horizontal pleiotropy, which occurs when a genetic variant affects the outcome through pathways other than the exposure of interest. Methods for detecting and adjusting for pleiotropy, including MR-Egger regression, are described. The review also discusses the performance of MR-Egger regression in practice, noting that it can provide a better fit to the data when heterogeneity is present. The review emphasizes the need for robust methods to account for heterogeneity and bias in MR analyses. It discusses various approaches, including weighted median and mode-based estimation, which are more robust to outliers and violations of assumptions. The review concludes that MR is a valuable tool for causal inference in epidemiology, and that further research is needed to improve the accuracy and reliability of MR analyses. The review also highlights the importance of standardized reporting guidelines to ensure the validity and reproducibility of MR studies.This review provides an overview of meta-analysis and Mendelian randomization (MR) in epidemiological research. MR uses genetic variants as instrumental variables to infer causal relationships between risk factors and health outcomes. Traditionally, MR combined results from separate studies using a small number of genetic variants, but recent advances allow the use of genome-wide association study (GWAS) summary data for large numbers of genetic variants. However, heterogeneity among causal estimates from different genetic variants may violate instrumental variable assumptions. The review explains the theoretical basis of MR, including instrumental variable theory, and describes methods for testing and adjusting for heterogeneity, such as random effects models, meta-regression, and robust regression. It also discusses the use of two-sample summary data MR, which allows the combination of GWAS summary data from different studies without requiring individual-level data. The review highlights the importance of addressing potential biases, such as horizontal pleiotropy, which occurs when a genetic variant affects the outcome through pathways other than the exposure of interest. Methods for detecting and adjusting for pleiotropy, including MR-Egger regression, are described. The review also discusses the performance of MR-Egger regression in practice, noting that it can provide a better fit to the data when heterogeneity is present. The review emphasizes the need for robust methods to account for heterogeneity and bias in MR analyses. It discusses various approaches, including weighted median and mode-based estimation, which are more robust to outliers and violations of assumptions. The review concludes that MR is a valuable tool for causal inference in epidemiology, and that further research is needed to improve the accuracy and reliability of MR analyses. The review also highlights the importance of standardized reporting guidelines to ensure the validity and reproducibility of MR studies.