The Mendelian Randomization Study Of Serum 25(OH)D Levels With The Muscular Development Among Children And Adolescents

ObjectiveIn order to understand the nutritional status of vitamin D and muscle development in children and adolescents aged 6-17 years,the related genetic variations in the 25（OH）D anabolic pathway were used as instrumental variable to perform Mendelian Randomization study,which was used to provide Causal inference evidence for the relationship between vitamin D and muscle development in children and adolescents.MethodsBased on 2019 Guangzhou Students Body Composition and Health Promotion Project,a stratified proportional random sampling method was used to select 4082 students aged 6 to 17 years old as the research objects.Collected body composition information such as grip strength,body fat,and lean mass through physical examination and body composition measurement.Questionnaire was used to investigate the intake of milk and dairy products,vitamin D supplements,sunscreen use,and exercise status.Blood samples were collected from students in fasting state and performed the serum 25（OH）D level detection and 25（OH）D related genotyping experiments.To analyze differences in physical characteristics and life behaviors of different genders,independent two sample t-test,Wilcocon rank sum test and independent sample² test were conducted.The Lambda-Median-Sigma method（LMS）was used to standardize the muscle development indexes（Grip,LMI,ALMI,ULMI,and LLMI）of the subjects by gender and age,then calculated the corresponding Z score and Constructed the percentile value and growth curves of muscle development level.Restricted cubic spline was used to test the nonlinear relationship between serum 25（OH）D levels and muscle development.Carried out general observational research through partial correlation analysis and multiple linear regression analysis by gender and development stages（childhood,adolescence）to understand the observational study association between serum 25（OH）D levels and muscle development in children and adolescents.Selected the common potential functional genetic variants of vitamin D metabolism-related genes through screening for publications of genome-wide association studies and bioinformatics analysis methods（rs12785878,rs1790349,rs4588,rs7041,rs1155563,rs10741657,rs12794714,rs6013897,rs2209314）.And performed instrumental variable core hypothesis tests（association,pleiotropy,and independence）on25（OH）D related genetic variations and genetic risk score（GRS）to construct instrumental variables.Generalized linear model（GLM）,Weighted Median Method and the inverse-variance weighting method（IVW）were applied to perform Mendelian randomization analysis on the relationship between 25（OH）D levels and muscle development in children and adolescents.Statistical analysis was performed by SAS 9.4 and R 3.6.3.Results1.Among 4082 children and adolescents aged 6-17,the average serum25（OH）D level was 69.96±20.13 nmol/L,and the 25（OH）D level of boys（73.29±21.06nmol/L）was significantly higher than girls（66.35±18.41nmol/L）（t=11.16,P<0.001）.Grip,LMI,ALMI,ULMI and LLMI of subjects all increased with age.Boys muscle development was higher than girls,and the changes are roughly same as the muscle development of Chinese children and adolescents.The development overall trend was as follows:boys grow slowly between the ages of 6 to 10 years old,and the10 to 14 years old age group is a period of rapid development,then growth slowing down at the age of 14 to 17 years old;girls grow slowly at the age of 6-9,then entering the period of rapid development from 9 to 13 years old,followed by slower development at the age of 12 to 17 years.2.General observational studies shown that 25（OH）D has no non-linear relationship with the muscle development of children and adolescents（all P>0.05）.Partial correlation analysis shown that 25（OH）D levels have significant positive correlations with grip Z score(r=0.10,P_FDR=0.001),LMI Z score(r=0.15,P_FDR<0.001),ALMI Z score(r=0.13,P_FDR<0.001),ULMI Z score(r=0.14,P_FDR<0.001)and LLMI Z score(r=0.12,P_FDR<0.001)in adolescent boys.In multiple linear regression analysis,increasing the 25（OH）D level can improve adolescent boys’grip Z score,LMI Z score,ALMI Z score,ULMI Z score and LLMI Z score(all P_FDR<0.001),and regression coefficient were 0.10,0.14,0.13,0.13,0.12respectively.No statistical correlation were found between the 25（OH）D level and the muscle development of child and adolescent girls(all P_FDR>0.05).3.Among the 9 genetic variants of the 4 target genes（DHCR7,GC,CYP2R1,CYP24A1）,rs12785878,rs1790349,rs4588,rs7041,rs1155563,rs10741657 and rs12794714 were statistically associated with the25（OH）D levels(all P_FDR<0.05),and rs4588 had pleiotropism with grip Z score(P_FDR=0.006).Unweighted GRS（modelⅠ）,weighted GRS（modelⅡ）and non-pleiotropic genetic variants weighted GRS（modelⅢ）were constructed by the above-mentioned genetic variants.Genetic variations and GRS models were in line with the core assumptions of instrumental variables,and there is no weak instrumental variable bias（all F>10）,so there was feasible to construct instrumental variables for Mendelian randomization analysis.4.In Mendelian randomization analysis,only rs12794714 was associated with ALMI Z score and LLMI Z score in adolescent girls(P_FDR<0.05).But other genetic variants and GRS models didn’t find any association between the 25（OH）D levels and the muscle development of children and adolescents in GLM,Weighted Median Method and IVW.Conclusions1.The vitamin D nutritional status of children and adolescents aged 6to 17 in Guangzhou is higher than the national average.However,vitamin D deficiency is still prevalent.2.The muscle development of children and adolescents in Guangzhou is consistent with the national level.Girls’muscle development is earlier than boys,but the development level is lower than boys.3.Based on the results of general observational studies and Mendelian randomization analysis,this study has not found that 25（OH）D levels can affect the muscle development of children and adolescents.