Advances in Exercise, Fitness, Performance Genomics in 2014
Advances in Exercise, Fitness, Performance Genomics in 2014
One study that examined whether cardiorespiratory fitness modified the polygenic risk for dyslipidemia was retained for this review. In this cross-sectional study, Tanisawa et al. measured serum levels of triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) and fitness using a maximal graded exercise test on a cycle ergometer in 170 Japanese men age 20–79 yr. Subjects were divided into low-fitness and high-fitness groups according to the reference V·O2max values from the prevention of lifestyle diseases, issued by the Ministry of Health, Labor, and Welfare of Japan (reference values in mL·kg·min: 39.0 for 20–39 yr, 35.0 for 40–59 yr, and 32.0 for 60 yr and older). The authors genotyped 19 SNP that met the following criteria: (1) showed genome-wide significant (P < 5 ×10) associations with TG, LDL-C, and/or HDL-C in any genomewide association study (GWAS) of individuals of European descent, (2) association was replicated (P < 0.05) in a Japanese population, and (3) minor allele frequency was more than 5% in the Japanese population. Three additive, weighted GRS were calculated based on SNP related to TG (seven SNP), LDL-C (five SNP), and HDL-C (nine SNP); and subjects were divided into tertiles (low, medium, and high) for each GRS for analysis.
In adjusted models, there was a significant interaction (Pinteraction = 0.028) between the TG-GRS group and the fitness group on TG levels. Triglyceride levels were 47 and 43 mm Hg higher in the high (P < 0.01) and middle (P < 0.05) TG-GRS groups, respectively, compared to the low TG-GRS group in the low-fitness group only, whereas no difference in TG levels was observed between the TG-GRS groups in the high-fitness group (Fig. 1).
(Enlarge Image)
Figure 1.
Associations among TG-GRS groups, fitness groups, and serum TG levels. Data shown are means (SD). Triglyceride was log transformed for analysis (data are shown as the original values). Data were analyzed by two-way analysis of covariance with adjustment for age, BMI, current or former smoking status, history of diabetes, alcohol consumption, and saturated fat intake. *P < 0.05 versus low-fitness subjects within the same GRS group. †P < 0.05 versus the low-GRS group within the same fitness group. ‡P < 0.01 versus the low-GRS group within the same fitness group. Adapted with permission from The American Physiological Society (31).
Furthermore, the number of individuals with hypertriglyceridemia (TG >= 150 mg·dL) was higher in the high and middle TG-GRS groups than in the low TG-GRS group in the low-fitness group only. Lastly, a significant interaction (Pinteraction < 0.05) between the TG-GRS and fitness was observed for body weight, as body weight was higher in the low-fitness group compared to the high-fitness group only in the high TG-GRS group. There was no interaction between the GRS group and the fitness group for LDL-C, HDL-C, or other lipoprotein-related traits (i.e., apolipoprotein B, apolipoprotein A-I, and oxidized LDL).
In summary, the study by Tanisawa et al. found that the polygenic risk for hypertriglyceridemia was attenuated by high fitness level. The study is strengthened by fitness (i.e., V·O2max) being directly measured and the inclusion of multiple GWAS-based SNP/loci. However, the study is limited by its small sample size and the reliance on a cross-sectional design. Thus, it is unknown whether intrinsic fitness, acquired fitness, or both is associated with the observed associations. It may be that high fitness intrinsically protects against hypertriglyceridemia regardless of TG-associated SNP. The authors suggest that future genetic association studies, including SNP associated with trainability of V·O2max, are needed to address this issue. Moreover, the trainability of TG and SNP associated with TG trainability could also play a role in the observed interaction. To date, no large-scale GWAS of lipid traits in a Japanese population has been performed. Thus, although the included white-based GWAS SNP were replicated (P < 0.05) in Japanese populations, given the differences in linkage disequilibrium among populations, the included SNP may not accurately represent the loci contributing the most to the genetic architecture of lipid traits in Japanese individuals. For example, a recent study in individuals of African American, East Asian, and European ancestry performed fine mapping of lipid GWAS loci and identified population-specific SNP that increased the trait variance explained.
It will be crucial to replicate and expand these findings in other Japanese cohorts and other ancestries. Further prospective and intervention studies are needed to examine the modification of polygenic effects on lipid traits by cardiorespiratory fitness. The study by Tanisawa et al. highlights the need for more studies that examine gene-physical activity, gene-exercise, or gene-fitness interactions on lipid traits at several loci simultaneously. A genomewide approach would allow for the identification of a panel of loci to be followed up in smaller, targeted replication studies.
Lipid and Lipoprotein Metabolism
One study that examined whether cardiorespiratory fitness modified the polygenic risk for dyslipidemia was retained for this review. In this cross-sectional study, Tanisawa et al. measured serum levels of triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) and fitness using a maximal graded exercise test on a cycle ergometer in 170 Japanese men age 20–79 yr. Subjects were divided into low-fitness and high-fitness groups according to the reference V·O2max values from the prevention of lifestyle diseases, issued by the Ministry of Health, Labor, and Welfare of Japan (reference values in mL·kg·min: 39.0 for 20–39 yr, 35.0 for 40–59 yr, and 32.0 for 60 yr and older). The authors genotyped 19 SNP that met the following criteria: (1) showed genome-wide significant (P < 5 ×10) associations with TG, LDL-C, and/or HDL-C in any genomewide association study (GWAS) of individuals of European descent, (2) association was replicated (P < 0.05) in a Japanese population, and (3) minor allele frequency was more than 5% in the Japanese population. Three additive, weighted GRS were calculated based on SNP related to TG (seven SNP), LDL-C (five SNP), and HDL-C (nine SNP); and subjects were divided into tertiles (low, medium, and high) for each GRS for analysis.
In adjusted models, there was a significant interaction (Pinteraction = 0.028) between the TG-GRS group and the fitness group on TG levels. Triglyceride levels were 47 and 43 mm Hg higher in the high (P < 0.01) and middle (P < 0.05) TG-GRS groups, respectively, compared to the low TG-GRS group in the low-fitness group only, whereas no difference in TG levels was observed between the TG-GRS groups in the high-fitness group (Fig. 1).
(Enlarge Image)
Figure 1.
Associations among TG-GRS groups, fitness groups, and serum TG levels. Data shown are means (SD). Triglyceride was log transformed for analysis (data are shown as the original values). Data were analyzed by two-way analysis of covariance with adjustment for age, BMI, current or former smoking status, history of diabetes, alcohol consumption, and saturated fat intake. *P < 0.05 versus low-fitness subjects within the same GRS group. †P < 0.05 versus the low-GRS group within the same fitness group. ‡P < 0.01 versus the low-GRS group within the same fitness group. Adapted with permission from The American Physiological Society (31).
Furthermore, the number of individuals with hypertriglyceridemia (TG >= 150 mg·dL) was higher in the high and middle TG-GRS groups than in the low TG-GRS group in the low-fitness group only. Lastly, a significant interaction (Pinteraction < 0.05) between the TG-GRS and fitness was observed for body weight, as body weight was higher in the low-fitness group compared to the high-fitness group only in the high TG-GRS group. There was no interaction between the GRS group and the fitness group for LDL-C, HDL-C, or other lipoprotein-related traits (i.e., apolipoprotein B, apolipoprotein A-I, and oxidized LDL).
In summary, the study by Tanisawa et al. found that the polygenic risk for hypertriglyceridemia was attenuated by high fitness level. The study is strengthened by fitness (i.e., V·O2max) being directly measured and the inclusion of multiple GWAS-based SNP/loci. However, the study is limited by its small sample size and the reliance on a cross-sectional design. Thus, it is unknown whether intrinsic fitness, acquired fitness, or both is associated with the observed associations. It may be that high fitness intrinsically protects against hypertriglyceridemia regardless of TG-associated SNP. The authors suggest that future genetic association studies, including SNP associated with trainability of V·O2max, are needed to address this issue. Moreover, the trainability of TG and SNP associated with TG trainability could also play a role in the observed interaction. To date, no large-scale GWAS of lipid traits in a Japanese population has been performed. Thus, although the included white-based GWAS SNP were replicated (P < 0.05) in Japanese populations, given the differences in linkage disequilibrium among populations, the included SNP may not accurately represent the loci contributing the most to the genetic architecture of lipid traits in Japanese individuals. For example, a recent study in individuals of African American, East Asian, and European ancestry performed fine mapping of lipid GWAS loci and identified population-specific SNP that increased the trait variance explained.
It will be crucial to replicate and expand these findings in other Japanese cohorts and other ancestries. Further prospective and intervention studies are needed to examine the modification of polygenic effects on lipid traits by cardiorespiratory fitness. The study by Tanisawa et al. highlights the need for more studies that examine gene-physical activity, gene-exercise, or gene-fitness interactions on lipid traits at several loci simultaneously. A genomewide approach would allow for the identification of a panel of loci to be followed up in smaller, targeted replication studies.
Source...