Kangjin Kim1, Yunhwan Lee2, Sungho Won3,4,5. 1. Institute of Health and Environment, Seoul National University, Seoul, South Korea. 2. Department of Public Health Sciences, Seoul National University, 1 Kwanak-ro Kwanak-gu, Seoul, 151-742, South Korea. 3. Institute of Health and Environment, Seoul National University, Seoul, South Korea. won1@snu.ac.kr. 4. Department of Public Health Sciences, Seoul National University, 1 Kwanak-ro Kwanak-gu, Seoul, 151-742, South Korea. won1@snu.ac.kr. 5. Interdisciplinary Program for Bioinformatics, College of Natural Science, Seoul National University, Seoul, South Korea. won1@snu.ac.kr.
Abstract
BACKGROUND: Metabolic syndrome is as a well-known risk factor for cardiovascular disease, which is associated with both genetic and environmental factors. Recently, the microbiome composition has been shown to affect the development of metabolic syndrome. Thus, it is expected that the complex interplay among host genetics, the microbiome, and environmental factors could affect metabolic syndrome. OBJECTIVE: To evaluate the relative contributions of genetic, microbiome, and environmental factors to metabolic syndrome using statistical approaches. METHODS: Data from the prospective Korean Association REsource project cohort (N = 8476) were used in this study, including single-nucleotide polymorphisms, phenotypes and lifestyle factors, and the urine-derived microbial composition. The effect of each data source on metabolic phenotypes was evaluated using a heritability estimation approach and a prediction model separately. We further experimented with various types of metagenomic relationship matrices to estimate the phenotypic variance explained by the microbiome. RESULTS: With the heritability estimation, five of the 11 metabolic phenotypes were significantly associated with metagenome-wide similarity. We found significant heritability for fasting glucose (4.8%), high-density lipoprotein cholesterol (4.9%), waist-hip ratio (7.7%), and waist circumference (5.6%). Microbiome compositions provided more accurate estimations than genetic factors for the same sample size. In the prediction model, the contribution of each source to the prediction accuracy varied for each phenotype. CONCLUSION: The effects of host genetics, the metagenome, and environmental factors on metabolic syndrome were minimal. Our statistical analysis suffers from a small sample size, and the measurement error is expected to be substantial. Further analysis is necessary to quantify the effects with better accuracy.
BACKGROUND: Metabolic syndrome is as a well-known risk factor for cardiovascular disease, which is associated with both genetic and environmental factors. Recently, the microbiome composition has been shown to affect the development of metabolic syndrome. Thus, it is expected that the complex interplay among host genetics, the microbiome, and environmental factors could affect metabolic syndrome. OBJECTIVE: To evaluate the relative contributions of genetic, microbiome, and environmental factors to metabolic syndrome using statistical approaches. METHODS: Data from the prospective Korean Association REsource project cohort (N = 8476) were used in this study, including single-nucleotide polymorphisms, phenotypes and lifestyle factors, and the urine-derived microbial composition. The effect of each data source on metabolic phenotypes was evaluated using a heritability estimation approach and a prediction model separately. We further experimented with various types of metagenomic relationship matrices to estimate the phenotypic variance explained by the microbiome. RESULTS: With the heritability estimation, five of the 11 metabolic phenotypes were significantly associated with metagenome-wide similarity. We found significant heritability for fasting glucose (4.8%), high-density lipoprotein cholesterol (4.9%), waist-hip ratio (7.7%), and waist circumference (5.6%). Microbiome compositions provided more accurate estimations than genetic factors for the same sample size. In the prediction model, the contribution of each source to the prediction accuracy varied for each phenotype. CONCLUSION: The effects of host genetics, the metagenome, and environmental factors on metabolic syndrome were minimal. Our statistical analysis suffers from a small sample size, and the measurement error is expected to be substantial. Further analysis is necessary to quantify the effects with better accuracy.