Fatty acid metabolism and SNPs related to obesity in a longitudinal study of Swedish men: Predictors of metabolic syndrome

The prevalence of obesity has risen dramatically in the last decades and its comorbidities, such as metabolic syndrome (MetS) and cardiovascular disease, are the major cause of mortality worldwide. This study was a part of the Uppsala longitudinal study of men (ULSAM), conducted in men born in the 1920s living in Uppsala at the age of 50 (N = 2322) and followed-up at age 70 (N = 1221).The aims of the present study were two-fold. Firstly, to investigate the associations between the single nucleotide polymorphisms (SNPs), on the SCD1 (stearoyl-CoA desaturase) gene and SNPs that had previously been associated with body mass index (BMI) or obesity in genome-wide association studies, with variables reflecting metabolic health. The variables reflecting metabolic health were fatty acid (FA) variables, fasting blood glucose, abdominal skinfold thickness, lipids and lipoproteins, and blood pressure in the ULSAM cohort at age 50. Secondly, to investigate to what extent SNPs, that significantly are associated with metabolic variables at age 50, affected the risk of developing MetS in the ULSAM cohort at age 70.SNPs with minor allele frequencies of > 1% and Hardy–Weinberg equilibriums of P < 0.001 were included. Furthermore, only one SNP in each linkage disequilibrium was selected in the final models. Subjects with genotypic call rate of < 95% were removed. Before the prediction modelling, a purposeful hypothesis-driven approach was used for the predictor variable selection. Variables were prioritized in the possible order of effect on the outcome health status at age 70. Non-normal variables were transformed and SNPs were coded as 0 for non-carriers, 1 for heterozygotes, and 2 for homozygotes of the minor allele. Subjects who had MetS at the beginning of the study, subjects who were underweight, all but one of each sibling group, and subjects who were missing MetS diagnostic variables at age 70 were removed before conducting binary regression analysis. Significant associations of SNPs were mainly with enzymes and/or FAs in cholesteryl esters, and high-density lipoprotein cholesterol (HDL-C). Variables of the ULSAM cohort that significantly correlated with one or more SNPs at age 50 were 16:1n-7, 22:6n-3, SCD-16, delta-5-desaturase (D5D), elongase, and HDL-C. The D5D enzyme activity had the greatest number of significant correlations with SNPs. Unexpectedly, none of the SNPs, that were selected because of their associations with BMI, were significantly associated with BMI in the ULSAM cohort, nor were SNPs on the SCD1 gene significantly associated with SCD activity in the ULSAM cohort. In the final binary regression model that best predicted MetS, BDNF rs7103411 heterozygotes (OR = 1.95, P = 0.01), homozygotes for the minor allele of FTO rs1558902 (OR = 3.07, P = 0.003), and variables at age 50, where low activity of D5D (OR = 0.05, P = 0.02), high percentage of 20:5n-3 (OR = 15.5, P < 0.001), high fasting blood glucose (OR = 1.84, P = 0.01), high abdominal skinfold thickness (OR = 17.7, P < 0.001), low serum HDL-C (OR = 0.16, P < 0.001), and high apolipoprotein B (OR = 10.2, P < 0.001) significantly increased the risk of developing MetS at age 70 in the ULSAM cohort. The findings suggest that some SNPs, that have previously been associated with BMI, may have an impact on FA metabolism. This may lead to various metabolic dysregulations later in life, making these SNPs, specifically FTO rs1558902 and BDNF rs7103411, and FA metabolism (specifically D5D activity), a way to assess people early who are at risk of developing MetS later in life in order to encourage lifestyle changes. These SNPs and FA variables could be potential drug-targets in cases where lifestyle changes do not work.

Author dc:creator

• Harpa Óskarsdóttir 1992-

Contributors dc:contributor

• Háskóli Íslands

• Líf- og læknavísindi
• Offita
• Lífshættir

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/1946/40428

OAI identifier oai:identifier

oai:skemman.is:1946/40428