Original Paper J Nutrigenet Nutrigenomics 2009;2:243–250 DOI: 10.1159/000255636
Received: June 22, 2009 Accepted: October 26, 2009 Published online: April 15, 2010
Association of Fatty Acid Desaturase Gene Polymorphisms with Blood Lipid Essential Fatty Acids and Perinatal Depression among Canadian Women: A Pilot Study Lin Xie Sheila M. Innis
Department of Paediatrics, Child and Family Research Institute, University of British Columbia, Vancouver, B.C., Canada
Key Words Fatty acid desaturase gene ⴢ Single nucleotide polymorphisms ⴢ Arachidonic acid ⴢ Eicosapentaenoic acid ⴢ Docosahexaenoic acid ⴢ Perinatal depression
Abstract Aims: The FADS1/FADS2 gene cluster encodes ⌬-5 and ⌬-6 desaturase, rate-limiting enzymes in metabolism of linoleic (LA) to arachidonic (ARA) and ␣-linolenic to eicosapentaenoic and docosahexaenoic acid (DHA). Single nucleotide polymorphisms (SNPs) in FADS1/FADS2 contribute to variability in blood lipid fatty acids. Altered n–6 and n–3 fatty acids have been related to perinatal depression (PPD). Methods: We genotyped rs174553, rs99780, rs174575, and rs174583 in FADS1/FADS2, analyzed blood lipid fatty acids and assessed PPD risk as an Edinburgh Postnatal Depression Scale (EPDS) score 610 for 69 pregnant women. Results: 21, 12 and 15% women had an EPDS score 610 at 36 weeks’ gestation, 2 and 6 months postpartum, respectively. Quantitative trait analysis showed an association between rs174575 and PPD risk at 36 weeks’ gestation and 6 months postpartum. With haplotype ACCC (major alleles) for rs174553, rs99780, rs174575, rs174583, respectively, as reference, GTCT was positively associated with PPD risk at 36 weeks’ gestation, p = 0.028, and higher LA and lower ARA in plasma (p = 0.0001,
© 2010 S. Karger AG, Basel 1661–6499/09/0025–0243$26.00/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/jnn
p ! 0.0001) and RBC ethanolamine phospholipids (p = 0.007, p = 0.005). Conclusions: We show that SNPs in FADS1/ FADS2 are associated with higher blood lipid LA and lower ARA and PPD risk. Copyright © 2010 S. Karger AG, Basel
Introduction
Perinatal depressive disorders, including prenatal and postpartum depression (PPD), defined as depressive disorders occurring during pregnancy or postpartum, are common disorders estimated to affect 10–20% of women in Western nations [1–4]. PPD is believed to be increasingly prevalent, and is of concern for several reasons that include an associated increased risk of adverse pregnancy outcome, and poor maternal-infant attachment and interaction with further implications for poor infant cognitive and behavior development [1, 5, 6]. These concerns provide compelling reasons to better understand the environmental variables and predisposing genetic factors that alone or in combination contribute an increased risk of PPD. Diet has undergone major changes over the last century and is known to influence neural metabolism through multiple direct and indirect mechanisms. Prof. Sheila M. Innis Department of Paediatrics, Child and Family Research Institute 950 West 28th Avenue, Vancouver BC V5Z 4H4 (Canada) Tel. +1 604 875 2431, Fax +1 604 875 3597 E-Mail sinnis @ interchange.ubc.ca
Changes in dietary fat with an increase in dietary n–6 fatty acids, particularly linoleic acid (LA, 18:2–6), and in the dietary n–6/n–3 fatty acid balance have been suggested to be contributing factors in several diseases, including depressive and other neurological disorders [7–11]. The biological basis for a role of fatty acids in neural function stems from extensive data which show that arachidonic acid (ARA, 20:4n–6) and the n–3 eicosapentaenoic acid (EPA, 20: 5n–3) and docosahexaenoic acid (DHA, 22:6n–3) in membrane phospholipids influences the metabolism of several neurotransmitters, including serotonin, and eicosanoids and other important signaling molecules, such as anadamides in the brain [12–15]. Some but not all studies have reported lower blood lipid DHA, EPA and increased ARA/n–3 fatty acid ratios in subjects suffering from depression [16–23]. Similarly, information linking fish intake and n–3 fatty acids to PPD is conflicting [24–26]. Studies addressing potential benefits of supplemental EPA and DHA for symptoms of depression have also given inconsistent results [27–29], with some favoring oils rich in EPA and suggesting DHA should not be given [30]. Although a major source of DHA and ARA for the brain is considered to be uptake from plasma [31, 32], the brain has abundant ⌬-6 and ⌬-5 desaturases, enzymes that synthesize ARA and EPA from LA and ␣-linolenic acid (ALA, 18:3n–3), respectively [33–35]. Genetic variation in FADS2 and FADS1, which encode ⌬-6 and ⌬-5 desaturase, respectively, is known to be associated with differences in plasma and red blood cell (RBC) ARA and EPA [36–39], raising the possibility that single nucleotide polymorphisms (SNPs) in FADS2 and FADS1 may influence the risk of problems such as PPD. Recently, we reported that minor allele variants of common SNPs in FADS1/FADS2 are associated with higher n–6 fatty acid precursor/product ratios in plasma and RBC phospholipids among pregnant women [38]. In the present report, we extend our previous studies to address whether SNPs in FADS1/FADS2 and their haplotypes are associated with PPD in our cohort of pregnant women followed prospectively from 16 weeks’ gestation and without previously known risk factors for PPD.
without knowledge of their dietary intakes with the exception that women following a vegan diet, or using fish oil or other n–3 fatty acid supplements were excluded. Women with known metabolic disease, including diabetes, immune disorders, or communicable diseases were ineligible for participation. All the women in the present report were seen at 16 and 36 weeks’ gestation, and 2 and 6 months postpartum, with complete data for genotyping, assessment of PPD and blood lipid n–6 and n–3 fatty acids. None of the women in the present study took a supplement containing EPA or DHA during their pregnancy. All of the women delivered 1 healthy term gestation infant (37–42 weeks’ gestation), with no complications. Further details relating to the subjects have recently been reported [38]. The protocol and procedures were approved by the Committee for Ethical Review of Research Involving Human Subjects at the University of British Columbia, and the British Columbia’s Children’s and Women’s Hospital. All subjects provided written informed consent prior to participation. Dietary Assessments Sociodemographic data, including age, parity, and highest level of education, family income and ethnic background, were collected from each subject using confidential questionnaires [38, 40, 41]. Dietary assessments were conducted at 16 and 36 weeks’ gestation using an interview-administered food frequency questionnaire designed to collect information on fat and fatty acid intakes, and intakes were estimated by entering data from the food records into nutrient analysis software (Food Processor 11; ESHA Research, Salem, Oreg., USA). No significant differences in the dietary n–3 and n–6 fatty acids were found with stage of gestation, thus results are given for 36 weeks’ gestation consistent with the assessment of PPD. Assessment of Postpartum Depression The Edinburgh Postnatal Depression Scale (EPDS) is a 10item self-administered questionnaire, validated and widely used for screening and monitoring PPD in women [42–44]. The EPDS was explained to each subject by a trained clinical assistant, and completed at 36 weeks’ gestation, then again at 2 and 6 months postpartum. Each woman was asked to underline the response that she felt best reflected herself over the previous 7 days, using a 4-point scale ranging from ‘No, not at all’ to ‘Yes, most of the time’, to give a maximum possible score of 30. A cut-off score of 9 or 10 is usually used to identify ‘possible depression’ [43]. For the purpose of the present study, women who scored 10 or above were considered ‘possibly depressed’. Sample Collection, Fatty Acid Analyses and Genotyping Fasting venous blood was collected in the outpatient laboratory of the British Columbia’s Children’s Hospital at 16 weeks’ gestation. For analysis of fatty acids, RBCs were separated from plasma by centrifugation (2,000 g, 15 min, 4 ° C), washed by resuspension in saline, recentrifuged, and the procedure repeated twice to remove contaminating plasma. Plasma and erythrocytes were frozen at –70 ° C until analyzed. For analysis, total lipids were extracted from the erythrocytes and plasma, then the plasma phospholipids and RBC ethanolamine phosphoglycerides (EPG, including ethanolamine plasmalogen and phosphatidylethanolamine) were separated by HPLC, recovered and their fatty acids quantified by GLC [40, 45, 46].
Subjects and Methods The present study involved 69 pregnant women, mean 8 SD of 33.5 8 3.6 years of age with no known maternal or fetal complications, all of whom were participants in a larger prospective study relating maternal fatty acid status in gestation to infant development [40]. The women were enrolled at 16 weeks’ gestation
244
J Nutrigenet Nutrigenomics 2009;2:243–250
Xie /Innis
25
25
**
plasma phospholipid and RBC EPG of women at 16 weeks’ gestation classified by rs174553 genotype. * p ! 0.01; ** p !0.001. Major (A/A) (black bars) compared to minor (G/G) (white bars) allele homozygotes; grey bars = heterozygotes.
15 10
**
5
15 10
* 5
0
0 LA ARA Plasma pospholipid
Genomic DNA was extracted from whole blood using the QIAamp DNA blood Mini kit (Qiagen Inc., Valencia, Calif., USA) and genotyping was performed with TaqMan SNP Genotyping Assays (Applied Biosystems) using real-time PCR [38]. We analyzed six SNPs covering a 34-kb long genomic region of the FADS1/FADS2 gene cluster. However, SNPs rs174561 and rs498793 were not consistent with Hardy-Weinberg equilibrium and no further analyses were done. The other four SNPs, rs174553, rs99780, rs174575 and rs174583, were consistent with the HardyWeinberg equilibrium, thus quantitative trait was applied to assess the relationship between SNP alleles and haplotypes and PPD risk, defined as an EPDS score 610. FADS1 and FADS2 are localized head-to-head on chromosome 11 (11q12–q13.1) and show a high degree of homology with exon/intron organization and rs174553, rs99780, rs174575 and rs174583 present in a highly preserved single linkage disequilibrium (LD) block (Dⴕ 1 0.9) that covers a 34-kb long genomic region from position 61331734 (rs174553) to 61366326 (rs174583), in order rs174553, rs99780, rs174575, and rs1745843, with rs174553 within intron 5 of FADS1, and rs99780 and rs174575 and rs174583 within introns of FADS2 [35, 47]. Statistical Analysis Results were analyzed with SPSS 15.0 software. Results for fatty acids are expressed as means 8 SD g/100 g fatty acids. The relationship between PPD and SNP alleles or haplotypes were analyzed by Unphased Software (Version 3.0.12, Frank Dudbridge MRC Biostatistics Unit, Cambridge, UK), and quantitative trait analysis applied using AddVal, which shows the estimated additive genetic value for a particular haplotype relative to the reference, assuming a normally distributed trait and small deviations from the mean.
Results
At 36 weeks’ gestation, 21% of the women had a total EPDS score of 10 or higher, and 12 and 15% of the women had a total EPDS score of 10 or higher at 2 and 6 months postpartum, respectively. Analyses of the dietary intakes Fatty Acid Desaturases, n–6 and n–3 Fatty Acids and Perinatal Depression
*
20
g/100 g fatty acid
Fig. 1. LA (18:2n–6) and ARA (20:4n–6) in
g/100 g fatty acid
20
LA ARA RBC ethanolamine phosphoglycerides
showed the median (25–75th range) of intakes of the major n–6 and n–3 fatty acids, LA and ALA, were 10.7 (8– 14.0) and 1.3 (1.0–1.9) g/day, respectively, and 110 (80– 160), 76 (30–120) and 118 (35–225) mg/day for ARA, EPA and DHA, respectively. For the 4 SNPs, rs174553, rs99780, rs174575 and rs174583, the major-minor alleles are A-G, C-T, C-G, C-T, respectively. As previously described, the major alleles of rs174583 and rs99780 were present in the same women, explained by the strong LD of these two SNPs (Dⴕ = 1, r2 = 1) [38]. Similarly, 19 of the 20 women homozygote for the major alleles of rs174583 were homozygote for the major allele of rs174553, and 19 of the 20 women homozygote for the G allele of rs174553 were homozygote for minor alleles of rs174583, explained by the strong LD (Dⴕ = 0.97, r2 = 0.94) of rs174553 and 174583 [38]. As a result, statistical analysis to address the relationship of minor allele variants of rs174553, rs174583 and rs99780 and PPD gave identical results. Minor allele carriers of rs174553 had statistically significantly lower plasma phospholipid and RBC EPG ARA and higher LA than major allele carriers (fig. 1) with no statistically significant differences in individual n–3 fatty acids or in their ratios among major and minor allele carriers. The results for rs99780 and 174583 were the same. As a group, minor allele carriers of rs174575 also had statistically significantly lower plasma phospholipid ARA than major allele carriers [38]. Quantitative Trait Analysis between SNP Alleles and Haplotypes and PPD Risk The quantitative trait analysis showed a statistical association between SNP rs174575 alleles and PPD risk at both 36 weeks’ gestation and 6 months postpartum (table 1). With the major allele C as the reference allele, the G allele of rs174575 was inversely associated with PPD J Nutrigenet Nutrigenomics 2009;2:243–250
245
Table 1. Relationship between SNP alleles in FADS2 and FADS1 and risk of perinatal depression
SNP
rs174553 rs99780 rs174575 rs174583
Allele
MarFeq
A G C T C G C T
0.5 0.5 0.5 0.5 0.681 0.319 0.5 0.5
36 weeks’ PPD risk
2 months’ PPD risk
6 months’ PPD risk
AddVal
AddVal
AddVal
0 –0.024 0 –0.006 0 –0.103 0 –0.006
p 0.591 0.893 0.044 0.893
0 –0.131 0 –0.022 0 –0.055 0 –0.022
p 0.509
0 –0.069 0 –0.064 0 –0.091 0 –0.064
0.637 0.290 0.637
p 0.068 0.086 0.051 0.086
MarFeq: marker frequency; AddVal: additive value of the minor compared to the major allele; PPD: prenatal and postpartum depression.
Table 2. Relationship between SNP haplotypes in FADS2 and FADS1 and risk of perinatal depression at 36 weeks’ gestation
Haplotype
MarFreq
A-C-C-C A-T-G-T G-C-C-C G-T-C-T G-T-G-T
0.493 0.007 0.007 0.181 0.312
AddVal 0 0.292 –0.494 0.093 –0.097
2
p value
0.091 1.954 0.980 4.808 4.997
0.763 0.162 0.322 0.028 0.025
The haplotype SNP order is rs174553-rs99780-rs174575rs174583. MarFeq: marker frequency is the frequency of the haplotype distribution. With haplotype A-C-C-C, major alleles for all SNP, as the reference. AddVal: additive value, indicates additive risk of perinatal depression for paired comparisons contributed to by the haplotype.
risk at 36 weeks’ gestation (p = 0.0437) and 6 months postpartum (p = 0.051). We found no statistically significant association between SNP and PPD risk at 3 other SNP loci, 174553, 174583 or 99780, although at 6 months postpartum a strong trend with all p values !0.1 for an inverse association of the minor allele variants and PPD risk is notable, given the pilot nature of this study (table 1). Thus, because minor allele variants are associated with lower plasma and RBC ARA (fig. 1), these results are consistent with a relationship between lower plasma ARA and low PPD risk, although these are statistical associations and an interpretation of cause and effect cannot be made. Due to the low subject number in our study, we limited further analysis between SNP haplotypes and 246
J Nutrigenet Nutrigenomics 2009;2:243–250
Table 3. Relationship between SNP haplotypes in FADS2 and FADS1 and major n–6 and n–3 fatty acids in plasma phospholipids and RBC ethanolamine phospholglycerides at 16 weeks’ gestation
Fatty acid
Haplotype
18:2n–6 20:4n–6 22:4n–6 22:5n–6 18:3n–3 20:5n–3 22:5n–3 22:6n–3
G-T-C-T G-T-C-T G-T-C-T G-T-C-T G-T-C-T G-T-C-T G-T-C-T G-T-C-T
Plasma phospholipid
RBC EPG
AddVal
AddVal p
0.381 –1.255 –0.249 –0.577 0.322 –0.106 –0.450 –0.308
p 0.0001
