Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

FADS1 FADS2 Gene Cluster, PUFA Intake and Blood Lipids in Children: Results from the GINIplus and LISAplus Studies

  • Marie Standl,

    Affiliation Institute of Epidemiology I, Helmholtz Zentrum München – German Research Centre for Environmental Health, Neuherberg, Germany

  • Eva Lattka,

    Affiliation Research Unit of Molecular Epidemiology, Helmholtz Zentrum München – German Research Centre for Environmental Health, Neuherberg, Germany

  • Barbara Stach,

    Affiliation Faculty of Medicine, Institute of Laboratory Medicine, Clinical Chemistry, and Molecular Diagnostics, University Hospital Leipzig, Leipzig, Germany

  • Sibylle Koletzko,

    Affiliation University of Munich Medical Centre, Dr. von Hauner Children's Hospital, Munich, Germany

  • Carl-Peter Bauer,

    Affiliation Department of Pediatrics, Technical University of Munich, Munich, Germany

  • Andrea von Berg,

    Affiliation Department of Pediatrics, Marien-Hospital Wesel, Wesel, Germany

  • Dietrich Berdel,

    Affiliation Department of Pediatrics, Marien-Hospital Wesel, Wesel, Germany

  • Ursula Krämer,

    Affiliation IUF, Leibniz Institut für Umweltmedizinische Forschung at the University of Düsseldorf, Düsseldorf, Germany

  • Beate Schaaf,

    Affiliation Medical Practice for Pediatrics, Bad Honnef, Germany

  • Stefan Röder,

    Affiliation Department for Environmental Immunology, Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany

  • Olf Herbarth,

    Affiliation Faculty of Medicine, Environmental Medicine and Hygiene, University of Leipzig, Leipzig, Germany

  • Anette Buyken,

    Affiliation Research Institute of Child Nutrition, University of Bonn, Dortmund, Germany

  • Tim Drogies,

    Affiliation Faculty of Medicine, Institute of Laboratory Medicine, Clinical Chemistry, and Molecular Diagnostics, University Hospital Leipzig, Leipzig, Germany

  • Joachim Thiery,

    Affiliation Faculty of Medicine, Institute of Laboratory Medicine, Clinical Chemistry, and Molecular Diagnostics, University Hospital Leipzig, Leipzig, Germany

  • Berthold Koletzko,

    Affiliation University of Munich Medical Centre, Dr. von Hauner Children's Hospital, Munich, Germany

  • Joachim Heinrich ,

    heinrich@helmholtz-muenchen.de

    Affiliation Institute of Epidemiology I, Helmholtz Zentrum München – German Research Centre for Environmental Health, Neuherberg, Germany

  • for the GINIplus Study Group¶ ,

    Membership of the the GINIplus Study Group is provided in the Acknowledgments.

  •  [ ... ],
  • and LISAplus Study Group¶

    Membership of the LISAplus Study Group is provided in the Acknowledgments.

  • [ view all ]
  • [ view less ]

Abstract

Background

Elevated cholesterol levels in children can be a risk factor for cardiovascular diseases in later life. In adults, it has been shown that blood lipid levels are strongly influenced by polymorphisms in the fatty acid desaturase (FADS) gene cluster in addition to nutritional and other exogenous and endogenous determinants. Our aim was to investigate whether lipid levels are determined by the FADS genotype already in children and whether this association interacts with dietary intake of n-3 fatty acids.

Methods

The analysis was based on data of 2006 children from two German prospective birth cohort studies. Total cholesterol, HDL, LDL and triglycerides were measured at 10 years of age. Six single nucleotide polymorphisms (SNPs) of the FADS gene cluster were genotyped. Dietary n-3 fatty acid intake was assessed by food frequency questionnaire. Linear regression modeling was used to assess the association between lipid levels, n-3 fatty acid intake and FADS genotype.

Results

Individuals carrying the homozygous minor allele had lower levels of total cholesterol [means ratio (MR) ranging from 0.96 (p = 0.0093) to 0.98 (p = 0.2949), depending on SNPs] and LDL [MR between 0.94 (p = 0.0179) and 0.97 (p = 0.2963)] compared to homozygous major allele carriers. Carriers of the heterozygous allele showed lower HDL levels [β between −0.04 (p = 0.0074) to −0.01 (p = 0.3318)] and higher triglyceride levels [MR ranging from 1.06 (p = 0.0065) to 1.07 (p = 0.0028)] compared to homozygous major allele carriers. A higher n-3 PUFA intake was associated with higher concentrations of total cholesterol, LDL, HDL and lower triglyceride levels, but these associations did not interact with the FADS1 FADS2 genotype.

Conclusion

Total cholesterol, HDL, LDL and triglyceride concentrations may be influenced by the FADS1 FADS2 genotype already in 10 year old children. Genetically determined blood lipid levels during childhood might differentially predispose individuals to the development of cardiovascular diseases later in life.

Introduction

Lipid concentrations in blood are associated with cardiovascular diseases [1][4]. Elevated cholesterol levels during childhood and adolescence increases the risk for cardiovascular diseases in adulthood. Moreover treatments which effectively lower cholesterol levels early in life have been shown to prevent disease manifestation in later life [5][7].

Among other factors, dietary polyunsaturated fatty acid (PUFA) intake can attenuate high blood lipid concentrations [8][10].

In addition to dietary influences, recent genome wide association studies have identified several genetic loci that are associated with blood lipid levels in adults [11][15]. Among the top hits for the determination of lipid levels are polymorphisms in the fatty acid desaturase (FADS) gene cluster. Lower total cholesterol [11], low-density lipoprotein (LDL) [11][13], high-density lipoprotein (HDL) [13], [14] and higher triglyceride levels [14], [15] are all associated with the minor alleles of the tested FADS polymorphisms. The genes FADS1 and FADS2, which are both located in this gene cluster, encode the enzymes delta-5-desaturase and delta-6-desaturase, which are involved in the conversion of dietary n-3 and n-6 fatty acids to their longer chain metabolites [16]. Polymorphisms in these genes are associated with n-3 and n-6 fatty acid levels in blood and several tissues [17][23]. Carriers of the minor alleles exhibit increased levels of desaturase substrates and decreased levels of desaturase products. This trend may be the result of lower transcription levels or diminished enzyme conversion rates in individuals carrying the minor alleles [17].

The previously reported strong association of FADS polymorphisms with lipid levels in adults makes these genes good candidates for genetic association studies focused on lipid levels in children. Given the role of the gene products in the conversion of dietary fatty acids to biologically important longer chain polyunsaturated fatty acids, which in turn have been shown to influence lipid levels, interaction analysis between genetic polymorphisms and dietary intake of fatty acids for individual lipid levels is of interest.

In this study we investigated whether genetic variations of the FADS gene cluster already pre-determines total cholesterol, HDL, LDL and triglyceride levels in 10 year old children and whether genetic variation interacts with dietary intake of n-3 PUFAs to determine individual blood lipid levels.

Results

Information on lipid concentrations in blood and FADS1 FADS2 genotype was available for 2006 children (1288 (64%) children from the GINIplus study and 718 (36%) children from the LISAplus study). Information on n-3 PUFA intake was only available for 1697 of these children (1100 (65%) children from the GINIplus study and 597 (35%) children from the LISAplus study).

Basic characteristics of the study population are presented in Table 1. Total cholesterol, LDL and triglyceride concentrations were significantly higher in the GINIplus study, whereas HDL concentrations were higher in the LISAplus study.

The genotype and allele frequencies of the six SNPs which were included in the analysis are shown in Table 2. There was no difference in the allele frequency distribution between the GINIplus and LISAplus studies.

thumbnail
Table 2. Characteristics of the SNPs in the FADS gene cluster.

https://doi.org/10.1371/journal.pone.0037780.t002

Five of the six SNPs (rs174545, rs174546, rs174556, rs174561 and rs3834458) are in high LD with each other. For these five SNPs, the pairwise squared correlations r2 ranged from 0.83 to 0.99, and Lewontin's D' ranged from 0.98 to 1.

For rs174575, the linkage disequilibrium is lower. The pairwise correlation r2 for this SNP ranged from 0.49 to 0.66 and Lewontin's D' ranged from 0.77 to 0.96.

Median levels of total cholesterol, HDL, LDL and triglyceride concentrations, stratified by FADS genotype, are presented in Table 3. Homozygous minor allele carriers had lower levels of total cholesterol and LDL compared to homozygous or heterozygous major allele carriers. In contrast, triglyceride concentrations were higher in minor allele carriers compared to homozygous major allele carriers. After adjustment for multiple testing (αcorr = 0.025), these associations remained significant for triglyceride levels and LDL (rs174556 and rs174561).

thumbnail
Table 3. Median total cholesterol, HDL, LDL and triglyceride concentrations [mmol/L] in with 25%- and 75%-quantiles stratified by FADS genotype (A: major allele/ a: minor allele).

https://doi.org/10.1371/journal.pone.0037780.t003

The results of the linear regression models for total cholesterol, HDL, LDL and triglyceride concentrations, the FADS genotype and n-3 PUFA intake are presented in Table 4. Adjusted means ratios for total cholesterol, LDL and triglyceride concentrations and effect estimates for HDL are shown.

thumbnail
Table 4. Results of linear regression models for total cholesterol, HDL, LDL and triglyceride concentrations, FADS genotype (A: major allele/ a: minor allele, reference: homozygous major allele) and n-3 PUFA intake (per IQR increase, IQR (n-3 PUFA)  = 0.04 mg/MJ) adjusted for gender, study centre, age, BMI, fasting status and total dietary energy intake [MJ].

https://doi.org/10.1371/journal.pone.0037780.t004

The p-values describing the association between n-3 PUFA intake and elevated concentrations of total cholesterol [MR = 1.01 for all six SNPs (p-value ranging from 0.0288 to 0.0884)], LDL [MR from 1.01 (p = 0.1075) to 1.02 (p = 0.0411)], HDL [β = 0.02 for all six SNPs (p-value between 0.0126 and 0.0306)] and reduced triglyceride levels [MR between 0.97 (p = 0.0219) and 0.98 (p = 0.0841)] were statistically significant. However, after correcting for multiple testing, only five (out of six) SNPs for HDL and triglyceride levels and rs3834458 remained significant.

Additional analyses showed similar results for the n-3 PUFAs ALA, EPA, DPA and DHA, which were combined into the total dietary n-3 PUFA intake (data not shown).

There was no association between dietary n-6 PUFA intake and lipid levels (data not shown).

Homozygous minor allele carriers had decreased levels of total cholesterol [MR ranging from 0.96 (p = 0.0093) to 0.98 (p = 0.2949)] and LDL [MR between 0.94 (p = 0.0179) and 0.97 (p = 0.2963)] compared to homozygous major allele carriers. After correcting for multiple testing (αcorr = 0.025), the association remained significant for rs174556 and rs174561 and total cholesterol and LDL.

HDL concentrations were reduced in carriers of the heterozygous genotype compared to homozygous major allele carriers [β between −0.04 (p = 0.0074) to −0.01 (p = 0.3318)]. These associations remained significant for four SNPs after correction for multiple testing.

Individuals carrying the heterozygous genotype showed significantly increased triglyceride levels compared to homozygous major allele carriers [MR ranging from 1.06 (p = 0.0065) to 1.07 (p = 0.0028)]. These associations remained significant after correcting for multiple testing. Although homozygous minor allele carriers also showed increased triglyceride levels, these effects did not reach statistical significance. Additional analyses restricted to fasting blood samples did not show substantially different results and the magnitude of the association between FADS genotype, n-3 PUFA and lipid concentration was similar. However, this result did not reach statistical significance, likely because of the reduced sample size (Table S1).

For each model, including the FADS SNP increased the percentage of explained variance compared to the model without any SNPs (Table S2). The maximal increase in the explained variance was 0.71% for the total cholesterol model (1.98% to 2.69%), 1.28% for the LDL model (from 4.21% to 5.49%), 0.96% for the HDL model (from 7.83% to 8.79%) and 0.48% for the triglycerides model (from 12.38% to 12.86%). Additional analyses stratified by study (GINIplus and LISAplus) showed similar results, although again, the lack of statistical significance is likely attributable to the reduced sample size (Figure S1).

There was no significant interaction between n-3 PUFA intake and FADS1 FADS2 genotype for any of the tested lipid concentrations (data not shown).

Discussion

The present study investigated the association between n-3 PUFA intake and FADS genotype with total cholesterol, HDL, LDL and triglyceride concentrations in 10-year-old children from the GINIplus and LISAplus birth cohort studies.

Although not all associations were statistically significant after adjustment for multiple testing, in these children, a higher n-3 PUFA intake was associated with higher total cholesterol, HDL and LDL and lower triglyceride levels. These associations remained significant after adjustment for multiple testing for HDL (five out of six tested SNPs) and triglyceride concentrations (rs3834458).

Minor alleles of FADS1 and FADS2 SNPs were significantly associated with higher levels of triglycerides and lower levels of total cholesterol, HDL, and LDL levels. However, not all of these associations reached statistical significance after correcting for multiple testing.

Generally, our results on trends in children are in line with previously published GWA studies that report an association between the minor allele of the tested FADS variant with lower total cholesterol [11], LDL [11][13], HDL [13], [14] and higher triglyceride [14], [15] concentrations in adults. Although the sample size of our cohort is relatively large, failure to reach statistical significance for all SNPs after correcting for multiple testing may be due to a lack of statistical power and small effect sizes. Nevertheless, our effect sizes for total cholesterol and triglycerides are comparable to those reported for adults [11], [14]. Bokor et al. [24] investigated the association between FADS haplotypes and lipid levels in adolescents, and reported no associations between any of the haplotypes and total cholesterol, HDL or LDL. However, the haplotype carrying the minor allele of rs174546 was significantly associated with higher triglyceride levels and the effect size reported is similar to that observed in our study.

The lack of statistical significance with respect to triglyceride levels in homozygous minor allele carriers may be due to the small size of this group (7% to 12% of the complete sample).

The inclusion of the FADS SNPs to the model lead to an increase of explained variance ranging from 0.48% for the triglyceride model to 1.28% for the LDL model, which showed a similar level as reported by several studies [25].

Lipid levels in children are determinants for cardiovascular diseases in adulthood [5][7]. Analysis of the underlying causes for disturbances in lipid metabolism during childhood can contribute to the prevention of cardiovascular diseases later in life. Our study suggests that the FADS1 FADS2 gene cluster may influence lipid levels in early life.

The underlying causal biological mechanism between an individual's FADS genotype and their lipid concentrations is not entirely clear. It is likely that the composition of polyunsaturated fatty acids in human tissues, which has been shown to be highly associated with the FADS genotype [17], is the direct link between the observed associations. Tanaka et al. [26] presumed that higher concentrations of the precursor fatty acids in minor allele carriers may result in increased membrane fluidity, and thus, in lower LDL. In addition to altered membrane fluidity, differential concentrations of long-chain PUFAs (LC-PUFAs) may lead to a change in the activation of transcription factors such as peroxisome proliferator activating receptor alpha (PPARA). Endogenous LC-PUFAs are natural ligands of PPARA [27], whose activation has been shown to influence the expression of apo-lipoproteins (e.g. ApoAI, Apo-AII, and ApoCIII) and enzymes (lipoprotein lipase) that are involved in the metabolism of lipoprotein particles [28][31].

The effect of FADS genotypes on fatty acid levels, which are also influenced by nutrition, leads to further interesting questions. One of which is whether there is a concerted interaction effect between FADS genotypes and PUFA intake on lipid levels. To date, few studies have dealt with this issue. Lu et al. [32] reported an association between FADS genotype and lipid levels, but only in groups with a high n-3 or n-6 PUFA intake. This result suggests an interaction between genotype and fatty acid intake, although a complete interaction analysis was not performed in this study. In our study, a higher intake of n-3 PUFA was associated with higher total cholesterol, HDL and LDL levels and lower triglyceride concentrations, although these effects were not significant after adjustment for multiple testing. We did not find any association between lipid levels, FADS genotype and dietary n-6 PUFA intake (data not shown). In our interaction analysis, we did not find a modification of the effect of n-3 PUFA intake on lipid concentrations in blood by the FADS genotype.

Additionally, the number of multiple comparisons has to be considered. In our manuscript, 24 hypotheses are investigated (6 SNPs and 4 traits). A correction for multiple testing is necessary, if several independent hypotheses are tested simultaneously. In the present manuscript, neither the SNPs, nor traits are independent. A more stringent approach, which would account for the number of outcome variables, would lead to a corrected alpha level of 0.0063. Taking this alpha level as a basis, the association between individuals carrying the heterozygous genotype and triglyceride levels compared to homozygous major allele carriers remains significant for five out of six SNPs. However, the reported associations with each of the selected six SNPs with traits were similar, possibly due to high correlation. A chance finding would be obvious, if just one of the tested SNPs were highly statistical significant, where the others were not. This is not the case in our analyses. Therefore, we do not consider these results as chance finding.

Nevertheless, our results suggest that there is an effect of n-3 PUFA intake as well as an effect of FADS genotype on lipid levels, although these effects do not interact with each other. It is known, that the conversion rate from dietary PUFA intake to longer chain metabolites is depending on the FADS genotype [17]. Therefore, it might be possible that dietary n-3 PUFA intake as well as the FADS genotype have a linear influence on the endogenous PUFA levels. This would be in line with the results presented by Moltó-Puigmartí et al. [22]. They reported lower DHA proportions in plasma phospholipids in women carrying the homozygous minor allele, but the DHA proportions increased with higher intake of fatty fish to a similar extent for all genotypes.

To our knowledge, this is the first study which has examined the complex associations between genetics, diet, and lipid levels in 10 year old children.

In addition to its strength, our study also faces some limitations which must be considered. Although lipid levels were measured from blood samples, the dietary fatty acid intake was assessed by a FFQ. The FFQ used in the present study measured dietary intake over the past 12 months and was validated for the dietary n-3 PUFA intake. Alternatively, it might be interesting to examine the association of FADS genes, lipid concentrations and measured fatty acid blood levels. A further limitation of our study is the low proportion of fasting blood samples (18.4%). The presented results were adjusted for fasting status. Additionally, the magnitude of the association between FADS genotype, n-3 PUFA and lipid concentration was similar when the sample was restricted to those with information on fasting blood, although statistical significance was lacking due to a reduced sample size (Results for triglyceride levels are presented in Table S1).

In order to increase the power, the two independent studies GINIplus and LISAplus were analyzed together. Additional analyses stratified for the GINIplus and LISAplus studies showed consistent results (Figure S1), except for LDL, although significance was missed due to the reduced sample size. This investigation addressed a complex hypothesis and is based on a specific data situation: Dietary fatty acid intake was assessed using a FFQ, which was especially developed and validated for dietary fatty acid intake in this specific study population [33], blood lipid levels of total cholesterol, HDL, LDL and triglycerides were measured and genotyping of six variants of the FADS1 FADS2 gene cluster was performed. Nevertheless, the results in the two independent studies GINIplus and LISAplus, which are based on the same methodology, are similar and our results are comparable to those reported for adults.

Our study suggests that the FADS1 FADS2 gene cluster may affect lipid levels already in childhood. Although the explained variance is low, and can therefore not be used for prevention or prediction purposes, these results underline the hypothesis that there is a causal association between dietary n-3 PUFA intake and lipid levels in children and may help to identify the causal biological mechanism.

However, further studies are needed to investigate the long-term effects of the impact of dietary intervention on the development of cardiovascular diseases, while considering the influence of the FADS gene cluster.

Materials and Methods

Study population

Data from two ongoing German birth cohort studies were included in this investigation: the German LISAplus (Life-style Related Factors on the Immune System and the Development of Allergies in Childhood PLUS the influence of traffic emissions and genetics) and GINIplus (German Infant Nutritional Intervention PLUS environmental and genetic influences on allergy development) studies. LISAplus is a population based birth cohort study in which a total of 3097 neonates were recruited between 1997 and 1999 from the German cities of Munich, Leipzig, Wesel and Bad Honnef. The participants were not pre-selected based on family history of allergic diseases [34]. A total of 5991 mothers and their newborns were recruited from Munich and Wesel into the GINIplus study between September 1995 and June 1998. Infants with at least one allergic parent and/or sibling were allocated to the interventional study arm which investigated the effect of different hydrolysed formulas consumed during the first year of life on the development of allergies [35]. All children without a family history of allergic diseases and children whose parents did not give consent for the randomized clinical trial were allocated to the non-interventional arm. Given that the entire current GINIplus study is composed of both the interventional and non-interventional arm, this cohort is also population-based. Detailed descriptions of the LISAplus and GINIplus studies have been published elsewhere [34][36]).

In both studies only individuals with Caucasian German descent were included.

For this analysis, only data from the 10 year follow-up is used. During this follow-up, blood samples were collected and in a subset of almost 20% of the children fasting blood samples could be collected.

For both studies, approval by the local Ethics Committees (Bavarian Board of Physicians, University of Leipzig, Board of Physicians of North-Rhine-Westphalia) and written consent from participant's families were obtained.

Dietary n-3 PUFA intake

A food frequency questionnaire (FFQ) was developed to measure a child's usual food and nutrient intake during one year, and more specifically, to estimate energy, fatty acid and antioxidant intake at 10 years of age. The FFQ comprised a list of 82 food items accompanied by several questions about the preferred fat and energy content of products, preparation methods, diet and food preferences, buying habits and dietary supplement use. The consumption frequencies and portion size estimates were converted to average consumption in grams per day and linked to the German Nutrient Data Base, version II.3.1 [37]. The design of the FFQ, including the selection of the food item list, validation, and the calculation of food and nutrient intake is described in more detail by Stiegler et al. [33].

The intake of n-3 PUFA was calculated by summing a child's daily intake of α-linolenic acid (ALA, 18: 3n-3), eicosapentaenoic acid (EPA, 20: 5n-3), docosapentaenoic acid (DPA, 22: 5n-3) and docosahexaenoic acid (DHA, 22: 6n-3).

Genotyping

Six single nucleotide polymorphisms (SNPs) in the FADS1 FADS2 gene cluster (rs174545, rs174546, rs174556, rs174561, rs174575 and rs3834458) were genotyped. Five of these variants (rs174545, rs174546, rs174556, rs174561 and rs3834458) are in strong linkage disequilibrium (LD) with each other (r2>0.7, D'>0.9) [17], and were selected based on previous publications in adult populations [17], [38]. Additionally, we included the rs174575 SNP in order to obtain a better coverage of the FADS gene cluster. This SNP was selected based on a previous publication in children [39]. By applying the tagger server program (http://www.broadinstitute.org/mpg/tagger/) in combination with HapMap (http://hapmap.ncbi.nlm.nih.gov/), we were able to tag 27 additional SNPs between base pair positions 61234329 and 61372379 in the FADS1 FADS2 gene cluster using three of our original SNPs (rs174545, rs174546 and rs174556). The efficiency was 10.7 fold, however of these new 27 SNPs, rs174561 and rs3834458 could not be used in the analysis as these are not included in the HapMap database. Genotyping of SNPs was conducted using the iPLEX (Sequenom, San Diego, CA, USA) method by means of matrix assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS, Mass Array; Sequenom) in one laboratory, according to the manufacturer’s instructions. Standard genotyping quality control included 10% duplicate and negative samples. The genotyping discordance rate was below 0.3%.

Measurement of lipids

The measurement of serum lipids and lipoproteins was performed using homogenous enzymatic colorimetric methods according to the manufactures instructions (Roche Diagnostics GmbH Mannheim). All parameters and controls were analysed on a Modular Analytics System from Roche Diagnostics GmbH Mannheim.

External controls were used in accordance with the guidelines of the German Society of Clinical Chemistry and Laboratory Medicine.

Statistical analysis

Because of the skewed distribution, total cholesterol, LDL and triglyceride concentrations were naturally log-transformed. Afterwards, linear regression modelling was used to assess the association between log-transformed total cholesterol, HDL, log-transformed LDL and log-transformed triglyceride concentrations, n-3 PUFA intake and FADS genotype. Therefore, for HDL, the regression coefficient β (Estimate) with standard deviation (Sd) is shown.

For total cholesterol, LDL and triglyceride, the results are presented as means ratio (MR) with 95% confidence interval (95% CI). The MR can be easily calculated for lognormal distributed variables and is derived by applying the exponential function on the regression coefficient β (i.e. MR = exp (β)). It describes the ratio of the mean of the outcome variable in one group compared to the mean of the outcome variable in the reference group. The MR can be interpreted as percentage change in the mean of the outcome variable in one group compared to the reference group adjusted for confounder variables.

The dietary n-3 PUFA intake was included in the regression analyses as nutrient density (n-3 PUFA intake divided by total energy intake). Additionally, the results were adjusted for total energy intake [40]. In order to compare the effect size of high (75% quantile) and low (25% quantile) dietary n-3 PUFA intake, the influence of n-3 PUFA intake is given per interquartile range increase (IQR (n-3 PUFA)  = 0.04 mg/MJ).

All regression models were adjusted for gender, study centre (Munich, Wesel, Leipzig and Bad Honnef), age, BMI at 10 years of age and fasting status.

Moreover, in additional analyses, an interaction between n-3 PUFA and FADS genotype (reference: homozygous major allele), was included in the linear regression models in order to test whether the effect of dietary n-3 PUFA intake is modified by FADS variants.

Statistical significance was defined by a two-sided alpha level of 5%. We corrected for multiple testing according to Nyholt [41]. In brief, this method takes the correlation pattern between the SNPs into account and reduces the number of variables in a set to the effective number of variables and provides thereby an estimate of the number of independent tests.

The alpha level is divided by the number of effective loci (which was computed as two, based on the number of effective loci of the six SNPs in the FADS gene cluster), which yields a corrected two-sided alpha level of 0.025 (5%/2 = 2.5%).

Differences between the GINIplus and LISAplus studies were tested using Fisher's exact test or Wilcoxon rank sum test. The association between lipid concentrations and FADS genotype was tested using Kruskal-Wallis rank sum test, a nonparametric method to test whether the median of the lipid concentrations is different between the FADS genotypes.

Statistical analysis was performed using the statistical software R, version 2.13.1 (http://www.R-project.org) [42].

Supporting Information

Figure S1.

Results of linear regression models on total cholesterol, HDL, LDL and triglycerides stratified for the GINIplus and LISAplus studies. Presented are means ratios (total cholesterol, LDL and triglycerides) and effect estimates (HDL) of FADS genotype (A: major allele/ a: minor allele, reference: homozygous major allele) and n-3 PUFA intake (per IQR increase, IQR (n-3 PUFA)  = 0.04 mg/MJ). All models are adjusted for gender, study centre, age, BMI and total dietary energy intake [MJ]. a) Total cholesterol b) HDL c) LDL d) Triglycerides

https://doi.org/10.1371/journal.pone.0037780.s001

(DOC)

Table S1.

Results of linear regression models restricted to fasting blood samples for triglyceride concentrations, FADS genotype (A: major allele/ a: minor allele, reference: homozygous major allele) and n-3 PUFA intake (per IQR increase, IQR (n-3 PUFA)  = 0.04 mg/MJ) adjusted for gender, study centre, age, BMI and total dietary energy intake [MJ].

https://doi.org/10.1371/journal.pone.0037780.s002

(DOC)

Table S2.

Percentage of variance explained in the models without and with the FADS variants.

https://doi.org/10.1371/journal.pone.0037780.s003

(DOC)

Acknowledgments

The GINIplus study group.

Helmholtz Zentrum München, German Research Center for Environmental Health, Institute of Epidemiology I, Munich (Heinrich J, Wichmann HE, Sausenthaler S, Zutavern A, Chen, Chih-Mei, Schnappinger M, Rzehak P); Department of Pediatrics, Marien-Hospital, Wesel (Berdel D, von Berg A, Beckmann C, Groß I); Department of Pediatrics, Ludwig Maximilians University, Munich (Koletzko S, Reinhardt D, Krauss-Etschmann S); Department of Pediatrics, Technical University, Munich (Bauer CP, Brockow I, Grübl A, Hoffmann U); IUF – Leibniz Research Institute for Environmental Medicine, Düsseldorf (Krämer U, Link E, Cramer C); Centre for Allergy and Environment, Technical University, Munich (Behrendt H).

The LISAplus study group.

Helmholtz Zentrum München, German Research Center for Environmental Health, Institute of Epidemiology I, Munich (Heinrich J, Wichmann HE, Sausenthaler S, Chen CM, Schnappinger M); Department of Pediatrics, Municipal Hospital “St. Georg”, Leipzig (Borte M, Diez U), Marien-Hospital Wesel, Department of Pediatrics, Wesel (von Berg A, Beckmann C, Groß I); Pediatric Practice, Bad Honnef (Schaaf B); Helmholtz Centre for Environmental Research – UFZ, Department of Environmental Immunology/Core Facility Studies, Leipzig (Lehmann I, Bauer M, Gräbsch C, Röder S, Schilde M); University of Leipzig, Institute of Hygiene and Environmental Medicine, Leipzig (Herbarth O, Dick C, Magnus J); IUF – Leibniz Research Institute for Environmental Medicine, Düsseldorf (Krämer U, Link E, Cramer C); Technical University Munich, Department of Pediatrics, Munich (Bauer CP, Hoffmann U); ZAUM – Center for Allergy and Environment, Technical University, Munich (Behrendt H, Grosch J, Martin F).

Author Contributions

Conceived and designed the experiments: SK CPB AvB DB UK B. Schaaf OH JH. Analyzed the data: MS. Wrote the paper: MS. Development of statistical analysis plan: EL JH. Lipid measurement: B. Stach TD JT. Interpretation of the findings: MS EL B. Stach SK CPB AvB DB UK B. Schaaf SR OH AB TD JT BK JH. Read and approved the manuscript: MS EL B. Stach SK CPB AvB DB UK B. Schaaf SR OH AB TD JT BK JH.

References

  1. 1. Kannel WB, Dawber TR, Kagan A, Revotskie N, Stokes J (1961) Factors of risk in the development of coronary heart disease – six year follow-up experience. The Framingham Study. Ann Intern Med 55: 33–50.
  2. 2. Miller NE, Miller GJ (1975) Letter: High-density lipoprotein and atherosclerosis. Lancet 1: 1033.
  3. 3. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR (1977) High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Am J Med 62: 707–714.
  4. 4. Kareinen A, Viitanen L, Halonen P, Lehto S, Laakso M (2001) Cardiovascular risk factors associated with insulin resistance cluster in families with early-onset coronary heart disease. Arterioscler Thromb Vasc Biol 21: 1346–1352.
  5. 5. Newman WP, Freedman DS, Voors AW, Gard PD, Srinivasan SR, et al. (1986) Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis. The Bogalusa Heart Study. N Engl J Med 314: 138–144.
  6. 6. Berenson GS, Srinivasan SR, Bao W, Newman WP, Tracy RE, et al. (1998) Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. The Bogalusa Heart Study. N Engl J Med 338: 1650–1656.
  7. 7. McGill HC, McMahan CA, Zieske AW, Sloop GD, Walcott JV, et al. (2000) Associations of coronary heart disease risk factors with the intermediate lesion of atherosclerosis in youth. The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Arterioscler Thromb Vasc Biol 20: 1998–2004.
  8. 8. Phillipson BE, Rothrock DW, Connor WE, Harris WS, Illingworth DR (1985) Reduction of plasma lipids, lipoproteins, and apoproteins by dietary fish oils in patients with hypertriglyceridemia. N Engl J Med 312: 1210–1216.
  9. 9. Calabresi L, Villa B, Canavesi M, Sirtori CR, James RW, et al. (2004) An omega-3 polyunsaturated fatty acid concentrate increases plasma high-density lipoprotein 2 cholesterol and paraoxonase levels in patients with familial combined hyperlipidemia. Metabolism 53: 153–158.
  10. 10. Mozaffarian D, Micha R, Wallace S (2010) Effects on coronary heart disease of increasing polyunsaturated fat in place of saturated fat: a systematic review and meta-analysis of randomized controlled trials. PLoS Med 7: e1000252.
  11. 11. Aulchenko YS, Ripatti S, Lindqvist I, Boomsma D, Heid IM, et al. (2009) Loci influencing lipid levels and coronary heart disease risk in 16 European population cohorts. Nat Genet 41: 47–55.
  12. 12. Sabatti C, Service SK, Hartikainen AL, Pouta A, Ripatti S, et al. (2009) Genome-wide association analysis of metabolic traits in a birth cohort from a founder population. Nat Genet 41: 35–46.
  13. 13. Chasman DI, Paré G, Mora S, Hopewell JC, Peloso G, et al. (2009) Forty-three loci associated with plasma lipoprotein size, concentration, and cholesterol content in genome-wide analysis. PLoS Genet 5: e1000730.
  14. 14. Kathiresan S, Willer CJ, Peloso GM, Demissie S, Musunuru K, et al. (2009) Common variants at 30 loci contribute to polygenic dyslipidemia. Nat Genet 41: 56–65.
  15. 15. De Silva NMG, Freathy RM, Palmer TM, Donnelly LA, Luan J, et al. (2011) Mendelian randomization studies do not support a role for raised circulating triglyceride levels influencing type 2 diabetes, glucose levels, or insulin resistance. Diabetes 60: 1008–1018.
  16. 16. Lattka E, Illig T, Koletzko B, Heinrich J (2010) Genetic variants of the FADS1 FADS2 gene cluster as related to essential fatty acid metabolism. Curr Opin Lipidol 21: 64–69.
  17. 17. Schaeffer L, Gohlke H, Müller M, Heid IM, Palmer LJ, et al. (2006) Common genetic variants of the FADS1 FADS2 gene cluster and their reconstructed haplotypes are associated with the fatty acid composition in phospholipids. Hum Mol Genet 15: 1745–1756.
  18. 18. Xie L, Innis SM (2008) Genetic variants of the FADS1 FADS2 gene cluster are associated with altered (n-6) and (n-3) essential fatty acids in plasma and erythrocyte phospholipids in women during pregnancy and in breast milk during lactation. J Nutr 138: 2222–2228.
  19. 19. Baylin A, Ruiz-Narvaez E, Kraft P, Campos H (2007) alpha-Linolenic acid, Delta6-desaturase gene polymorphism, and the risk of nonfatal myocardial infarction. Am J Clin Nutr 85: 554–560.
  20. 20. Malerba G, Schaeffer L, Xumerle L, Klopp N, Trabetti E, et al. (2008) SNPs of the FADS gene cluster are associated with polyunsaturated fatty acids in a cohort of patients with cardiovascular disease. Lipids 43: 289–299.
  21. 21. Lattka E, Rzehak P, Szabó É, Jakobik V, Weck M, et al. (2011) Genetic variants in the FADS gene cluster are associated with arachidonic acid concentrations of human breast milk at 1.5 and 6 mo postpartum and influence the course of milk dodecanoic, tetracosenoic, and trans-9-octadecenoic acid concentrations over the duration of lactation. Am J Clin Nutr 93: 382–391.
  22. 22. Moltó-Puigmartí C, Plat J, Mensink RP, Müller A, Jansen E, et al. (2010) FADS1 FADS2 gene variants modify the association between fish intake and the docosahexaenoic acid proportions in human milk. Am J Clin Nutr 91: 1368–1376.
  23. 23. Koletzko B, Lattka E, Zeilinger S, Illig T, Steer C (2011) Genetic variants of the fatty acid desaturase gene cluster predict amounts of red blood cell docosahexaenoic and other polyunsaturated fatty acids in pregnant women: findings from the Avon Longitudinal Study of Parents and Children. Am J Clin Nutr 93: 211–219.
  24. 24. Bokor S, Dumont J, Spinneker A, Gonzalez-Gross M, Nova E, et al. (2010) Single nucleotide polymorphisms in the FADS gene cluster are associated with delta-5 and delta-6 desaturase activities estimated by serum fatty acid ratios. J Lipid Res 51: 2325–2333.
  25. 25. Stranger BE, Stahl EA, Raj T (2011) Progress and promise of genome-wide association studies for human complex trait genetics. Genetics 187: 367–383.
  26. 26. Tanaka T, Shen J, Abecasis GR, Kisialiou A, Ordovas JM, et al. (2009) Genome-wide association study of plasma polyunsaturated fatty acids in the InCHIANTI Study. PLoS Genet 5: e1000338.
  27. 27. Fruchart JC, Duriez P, Staels B (1999) Peroxisome proliferator-activated receptor-alpha activators regulate genes governing lipoprotein metabolism, vascular inflammation and atherosclerosis. Curr Opin Lipidol 10: 245–257.
  28. 28. Hertz R, Bishara-Shieban J, Bar-Tana J (1995) Mode of action of peroxisome proliferators as hypolipidemic drugs. Suppression of apolipoprotein C-III. J Biol Chem 270: 13470–13475.
  29. 29. Schoonjans K, Peinado-Onsurbe J, Lefebvre AM, Heyman RA, Briggs M, et al. (1996) PPARalpha and PPARgamma activators direct a distinct tissue-specific transcriptional response via a PPRE in the lipoprotein lipase gene. EMBO J 15: 5336–5348.
  30. 30. Vu-Dac N, Schoonjans K, Laine B, Fruchart JC, Auwerx J, et al. (1994) Negative regulation of the human apolipoprotein A-I promoter by fibrates can be attenuated by the interaction of the peroxisome proliferator-activated receptor with its response element. J Biol Chem 269: 31012–31018.
  31. 31. Vu-Dac N, Schoonjans K, Kosykh V, Dallongeville J, Fruchart JC, et al. (1995) Fibrates increase human apolipoprotein A-II expression through activation of the peroxisome proliferator-activated receptor. J Clin Invest 96: 741–750.
  32. 32. Lu Y, Feskens EJ, Dollé ME, Imholz S, Verschuren WM, et al. (2010) Dietary n-3 and n-6 polyunsaturated fatty acid intake interacts with FADS1 genetic variation to affect total and HDL-cholesterol concentrations in the Doetinchem Cohort Study. Am J Clin Nutr 92: 258–265.
  33. 33. Stiegler P, Sausenthaler S, Buyken AE, Rzehak P, Czech D, et al. (2010) A new FFQ designed to measure the intake of fatty acids and antioxidants in children. Public Health Nutr 13: 38–46.
  34. 34. Heinrich J, Bolte G, Hölscher B, Douwes J, Lehmann I, et al. (2002) Allergens and endotoxin on mothers' mattresses and total immunoglobulin E in cord blood of neonates. Eur Respir J 20(3): 617–623.
  35. 35. von Berg A, Koletzko S, Grübl A, Filipiak-Pittroff B, Wichmann HE, et al. (2003) The effect of hydrolyzed cow's milk formula for allergy prevention in the first year of life: the German Infant Nutritional Intervention Study, a randomized double-blind trial. J Allergy Clin Immunol 111(3): 533–540.
  36. 36. von Berg A, Krämer U, Link E, Bollrath C, Heinrich J, et al. (2010) Impact of early feeding on childhood eczema: development after nutritional intervention compared with the natural course – the GINIplus study up to the age of 6 years. Clin Exp Allergy 40(4): 627–636.
  37. 37. Hartmann BM, Bell S, Vásquez-Caicedo AL, Götz A, Erhardt J, et al. (2005) Der Bundeslebensmittelschlüssel. German Nutrient Data Base. Karlsruhe.
  38. 38. Rzehak P, Heinrich J, Klopp N, Schaeffer L, Hoff S, et al. (2009) Evidence for an association between genetic variants of the fatty acid desaturase 1 fatty acid desaturase 2 (FADS1 FADS2) gene cluster and the fatty acid composition of erythrocyte membranes. Br J Nutr 101: 20–26.
  39. 39. Caspi A, Williams B, Kim-Cohen J, Craig IW, Milne BJ, et al. (2007) Moderation of breastfeeding effects on the IQ by genetic variation in fatty acid metabolism. Proc Natl Acad Sci U S A 104: 18860–18865.
  40. 40. Willett WC, Howe GR, Kushi LH (1997) Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 65: 1220S–1228S; discussion 1229S–1231S:
  41. 41. Nyholt DR (2004) A simple correction for multiple testing for single-nucleotide polymorphisms in linkage disequilibrium with each other. Am J Hum Genet 74: 765–769.
  42. 42. R Development Core Team (2011) R: A Language and Environment for Statistical Computing. Vienna, Austria.