Changes in Plasma Lipid Homeostasis Observed in Chronic Fatigue Syndrome Patients

A study of 60 CDC-defined chronic fatigue syndrome (CFS) patients and 39 age- and sex-matched controls was undertaken to determine whether there were any alterations in plasma lipid profiles from CFS patients compared with control subjects. Plasma samples were taken from fasted (12 h) subjects and the lipid components were extracted, saponified and analyzed by capillary gas chromatography-mass spectrometry (GC-MS). It was found that CFS patients had different plasma saponified lipid profiles compared with control subjects (standard discriminant function analysis p <0.02). The first and second most important factors, discriminating the CFS patients from the controls, were a decrease in trans-9-octadecenoic acid and an increase in octadecenoic acid, respectively. The relative abundance of trans-9-octadecenoic acid in the CFS patients was 0.74% compared with 0.99% in the controls (p < 0.02) and the trans-9-octadecenoic acid: octadecenoic acid ratio was 0.15 in the CFS patients compared with 0.27 in controls (p < 0.00002). The younger CFS patients ( 24 years old) on the basis of their lipid profiles, whereas no such differentiation could be made within the control group. These results support an organic basis to CFS and indicate that compliance with the clinical definition of CFS does not necessarily produce a homogeneous group of patients. The differential age-related changes may occur as a result of CFS or may actually represent predisposing factors leading to and sustaining CFS.

Keywords: chronic fatigue syndrome, plasma lipids, trans-9-octadecenoic acid, octadecenoic acid, fatigue aetiology, fatty acids.


Chronic fatigue syndrome (CFS) is characterized by unexplained, persistent or relapsing, chronic and debilitating fatigue lasting six or more consecutive months [ 1]. The diagnosis requires the exclusion of other known fatigue-related diseases and compliance with a clinical definition [ 1]. No single aetiological agent has yet been found but this may be the result of the heterogeneity of the patient groups studied under the guise of the current definitions. Some investigators have suggested that CFS results from a common host response to a heterogeneous group of aetiological agents involving immune and neural function. The extensive range of symptoms relating to fatigue and pain reported by CFS patients suggests alterations in metabolism and homeostasis, and an extensive evaluation of the molecular basis of the syndrome and its potential subgroupings is therefore required. Changes in serum and cerebrospinal fluid monoamines [ 2], as well as an alteration in the hypothalamic-pituitary-adrenal axis, have been reported in American cohorts of CFS patients [ 3]. Significant anomalies in urinary excretion strongly associated with symptom incidence [ 4] have also been noted in an Australian cohort of CFS patients. These molecular data may prove useful in the future development of objective diagnostic criteria and may also allow differentiation between various subgroups of this potentially heterogeneous group of subjects.

Recent evidence indicates that 2-5A synthetase RNAase L activity is upregulated in patients with CFS [ 5], suggesting ongoing viral infections or autoimmune disease. In this context, treatment with the antiviral agent Poly( 1)-Poly(C12U) results in improvement of the symptoms of CFS subjects [ 6]. Although the literature does not support a common viral aetiology (reviewed in [ 7]), these data nevertheless suggest that viral infections may be implicated in CFS. Viral infections can alter essential fatty acid metabolism leading to significant reductions in n-6 essential fatty acids in cell membranes [ 8]. A substantial improvement, particularly in the expression of myalgia, was reported in 85% of post-viral fatigue syndrome patients following supplementation with essential fatty acids [ 8]. Viral infections may induce changes in fatty acid metabolism which thereby alter the host response leading, to and sustaining, CFS.

If patients with CFS have alterations in lipid homeostasis, it should be possible to detect these alterations in blood plasma samples (12 h fasted) representing a biochemical "snapshot" of lipid homeostasis with minimal dietary input. To test this hypothesis, we screened saponified plasma lipid products of CFS patients and non-CFS control subjects using capillary gas chromatography-mass spectrometry (GC-MS). The aim of this study was to assess whether qualitative or quantitative changes in plasma lipid profiles occur in CFS patients compared with control subjects.


Study Subjects
Patients were examined and diagnosed by two clinicians (M.D. and P.D.) from August 1992 to February 1994. CFS study patients (CFS group, n = 60, range 11-54 years) and non-CFS subjects (control group, n - 39, range 17-54 years) were identified according to the Centers for Disease Control (CDC, Atlanta, USA) CFS research criteria [ 9]. The 39 non-fatigued control subjects were recruited from healthy relatives of the patients and from unrelated volunteer subjects. No differences were found in any of the age and sex characteristics between the 2 study groups (CFS, age +/- SD = 35.7 +/- 10.6 years, 62% female; Control, age +/- SD = 34.5+/- 10.4, 54% female). All subjects had given their informed consent and the project was approved by the Human Research Ethics Committee, the University of Newcastle.

Lipid Extraction of Plasma Samples
Whole blood (10 mL) was collected from fasted patients and control subjects (12 h) by venepuncture into a lithium heparin vacutainer (Becton Dickinson) at either the Environmental Medical Centre (Sydney, Australia) or the University of Newcastle (Australia). The collection of samples from patients and controls occurred on a regular basis over the sampling period and samples were processed immediately on receipt. The samples were chilled on ice and transported to the laboratory. Plasma samples were coded and then processed blind by the laboratory staff. Plasma was separated by centrifugation at 1600 g for 25 min at 4 Celsius. Aliquots of plasma (1 mL) were transferred into 12 mL PTFE-lined screw-capped glass extraction tubes and butylated-hydroxy-toluene (BHT) was added as an antioxidant (100 Mu L, 10 gL-1 in methanol) [ 10]. Samples were stored at -20 Celsius. Extraction was performed within 48 h after separation.

Lipophilic components were extracted from the plasma (1 mL) by sequentially adding 2 mL methanol, 2 mL hexane and 2 mL diethyl ether. Samples were vortexed for 60 s and then centrifuged at 1600 g for 10 min at 10 Celsius in a Fisons MSE Chilspin centrifuge. The upper hexane/ether phase was transferred to a 50 mL glass pear-shaped flask. The aqueous residue was re-extracted twice more with 4 mL fresh hexane/diethyl ether (1:1). The combined hexane/ether extracts were evaporated using rotary evaporation under vacuum until approximately 1 mL of solvent remained. This extract was transferred to a PTFE-lined screw-capped glass derivatization tube with 3 x 1 mL hexane/diethyl ether (1:1) washes. The solvent was removed from the derivatization tube using a stream of high purity nitrogen gas until approximately 1 mL of the solvent remained and the sample was stored at -20 Celsius. The sample was taken to dryness just prior to saponification which was performed within 24 h of extraction. This method resulted in better extraction recoveries and reproducibility compared with other reported protocols [ 11-13] and was more efficient in the removal of aqueous components.

The complex lipids in the samples were saponified by incubating the dried lipid extract with 100 gL 10% KOH in 95% (v/v) ethanol at 100 Celsius for 2 h [ 14]. At the end of this period the sample was acidified with 60 Mu L 6 M HCI [ 15-17]. The sample was snap frozen using liquid nitrogen and freeze-dried for at least 30 min. Once dry, 500 Mu l 95% (v/v) ethanol was added and the sample was mixed by vortex for 20 s and centrifuged at 2000 rpm for 2 min in a Sorvall bench centrifuge. The ethanol fraction was transferred to a clean derivatization tube and two further aliquots of alcohol were used to extract the remaining saponified lipids. The combined ethanol extracts were allowed to dry on the freeze-drier for 2 h before derivatization was performed. The derivatized components were stored at 4 Celsius overnight before analysis by GC-MS.

Following saponification the dry lipid samples were derivatized by adding 50 Mu L 2% (w/v) methoxyamine hydrochloride in pyridine (MOX) and incubated at 60 Celsius for 30 min. Bis(trimethylsilyl)trifluoro-acetamide (BSTFA, 100 Mu L) was added and the sample was heated at 100 Celsius for a further 60 min to form the trimethylsilyl derivatives (TMS) [ 12]. This method of derivatization was selected to allow the evaluation of sterols and the non-fatty acid components of complex lipids simultaneously with the fatty acids. Prior to injection into the gas chromatograph (GC), each sample was reheated at 100 Celsius for 10 min, cooled to room temperature and the solvent was removed using nitrogen gas. Fresh BSTFA (100 Mu L) was added and the derivatives were heated for a further 5 min at 100 Celsius to ensure that the entire sample was solubilized in the BSTFA. Each sample was centrifuged at 2000 rpm for 3 min in a Sorvall bench centrifuge prior to injection (0.5 Mu L) into the GC-MS.

Gas Chromatography-Mass Spectrometry
Derivatized sample components were separated using a Hewlett Packard 5890 series II GC and detected using a Hewlett Packard series 5971A Mass Selective Detector (MSD). The data were stored and processed on a Hewlett Packard Unix-based Chemstation. The GC was fitted with a 25 m, 0.20 mm internal diameter HP1 fused silica capillary column (film 0.33 Mu m) and a split/splitless injector. The column gave good separation of the cis and trans isomers of the unsaturated fatty acids. The GC was run with the following settings: injector temperature 300 Celsius, temperature program 80-300 Celsius at 3 Celsius min-1, with a 2 min hold at 80 Celsius and a 10 min hold at 300 Celsius. The MSD operated in El mode and was set to scan from 45-650 atomic mass units every second. Where possible, the peaks were identified by HP-UX Chemstation computer search of user-generated reference libraries (incorporating retention indices and mass spectra) and the WILEY Database Trademark. Lipid profile patterns were generated by assessing each peak as a percentage abundance of the total peak area of all the peaks in the chromatogram. Single ion monitoring (SIM) was utilized for the analysis of cis-9, 12-C18:2 (m/z = 337) and cis/trans 9-C18:1 (m/z = 339) to achieve full resolution of these lipid components. The fatty acid and sterol data were converted to mmol L-1 using available standards. Where standards were unavailable, or the compound could not be fully identified, standard curves were extrapolated from the closest reference standard (e.g. linoleic acid for C18:2(a) and (b)). The data were also expressed as relative abundance values where each peak was calculated as a percentage of the total area of all the assessed peaks. Lipid profile patterns were evaluated as either a relative abundance (RA) profile or a peak area (PA) profile (and expressed as mmol L-1). The fatty acid concentrations measured by this broader-based extraction and TMS-derivatization technique (which included assessment of sterols and other saponification products) generally corresponded to those ranges reported for more specific fatty acid analysis by methyl esters as shown in Table 1. The levels of DGLA and arachidonic acid were slightly higher than the reported reference range which may reflect the use of plasma instead of serum, or alternatively an improved extraction recovery/preservation by this method, or it may simply represent a variation within the Australian population mediated by diet.

Statistical Analysis
The percentage composition values for the plasma lipid components were arcsine transformed while the concentration/peak area values were log transformed prior to statistical analysis to give normal distributions and allow parametric statistical analysis. Group differences were assessed by standard discriminant function and multiple regression analyses as appropriate. Student's t-tests, Duncan's multiple range and Chi-square probability tests were used where indicated. These data were processed using Access Trademark (Ver. 97, Microsoft), Excel Trademark (Ver. 97, Microsoft) and Statistic Trademark (Ver. 5. 97 Edition, Statsoft, Tulsa).

Bis(trimethylsilyl)trifluoro-acetamide, methoxyamine hydrochloride, BHT and the fatty acid standards were obtained from Sigma Chemical Company. All other chemicals were obtained from BDH. All reagents used were of analytical grade quality (AnalaR) except the hexane and methanol which were Pesticide Analysis grade.


Saponified Lipids
The saponified lipid products were detected as TMS derivatives by GC-MS in the plasma samples and the concentration/relative abundance data are summarized in Table 2. The fatty acid components included saturated fatty acids (12:0 to 24:0), n-7 and n-9 monounsaturated fatty acids, and n-6 and n-3 polyunsaturated fatty acids. A number of polar saponification components were detected (such as phosphate, 1-and 2-glycerophosphate and glycerol) which were presumably derived from the alkaline hydrolysis of cholesterol esters, phospholipids and triacylglycerides. Five additional fatty acids were detected and are presented in Table 2 as "unassigned fatty acids". Although the chain length of these fatty acids and their level of unsaturation could be determined from their mass spectra, the position of unsaturation and the isomerism could not be determined by comparison with available reference standards. Cholesterol was the most abundant lipid saponification product measured in the plasma. Small quantities of the cholesterol precursor 5 Alpha-cholest-7-en-3 Beta-ol (lathosterol), as well as the plant sterols, campest-5-en-3 Beta-ol (campesterol) and sitost-5-en-3 Beta-ol (sitosterol), were also detected in the plasma samples. The molar concentrations (mmol L-1) for the fatty acids and sterols are also shown in Table 2, and these were consistent with the known reference ranges [ 18]. The levels measured for oleic acid were slightly lower, and the levels measured for dihomo-Gamma-linolenic acid (DGLA) and arachidonic acid were slightly higher than the range reported in the literature which may reflect population differences and/or the use of plasma instead of serum.

CFS vs Controls
The array of saponified lipid products measured in plasma samples provides a complex profile (Table 2) which can be assessed by multivariate analysis to compare lipid metabolism in the CFS patients with that in control subjects. Analysis of the RA data by standard discriminant function analysis indicated that the plasma lipid profiles from CFS patients were significantly (p < 0.02) different compared with those from controls (Table 3a). Forward stepwise discriminant analysis revealed that the first and second most important lipids for discriminating the CFS patients from the controls were trans-9-octadecenoic acid and octadecenoic acid, respectively, using either the RA or PA data. There was no detectable trans-9-octadecenoic acid in the plasma of 15 (25%) of the 60 CFS patients, whereas trans-9-octadecenoic acid was absent in only one (2.6%) of the 39 control subjects (p < 0.003). When the standard discriminant function model was used to classify individual subject profiles as either CFS or control on the basis of their lipid profile, it was able to assign subjects into their respective clinical groups with 87% accuracy (Table 3a). The same model was used to assign an additional 34 CFS patients (average age 51 +/- 15, range 23-97) and 5 control subjects (average age 31-41, range 23-47), which were not used in the study, to test the power of the predictive model. The discriminant function was able to assign these external subjects with a 67% accuracy.

Since trans-9-octadecenoic acid (reduced in CFS) and octadecenoic acid (increased in CFS) were the major discriminant lipids for differentiating CFS patients and controls, the trans-9-octadecenoic acid:octadecenoic acid ratio (PA data) was also calculated and compared between the two groups. The trans-9-octadecenoic acid:octadecenoic acid ratio in the CFS group (mean +/- SD, 0.15 +/- 0.11) was lower (p <0.000004) than that measured in the control group (0.27 +/- 0.15).

Heterogeneity in the CFS Group
Standard logistic regression analysis indicated that age was not correlated with changes in the RA lipid profiles in either the control group or the CFS group. Similar analysis of the PA lipid data in the control group also showed no correlation. In contrast, however, the PA lipid data for the CFS group were significantly correlated with age (R2 = 0.76, p < 0.035), where cis-15-tetracosenoic acid (nervonic acid, positive association), cis-7-tetradecenoic acid and trans-9-octadecenoic acid (negative associations) were the primary correlates in the regression model. These results suggest that significant alterations in lipid homeostasis occur with age in the CFS group, but not in the control group.

To assess whether the age dependent associations in lipid homeostasis lead to heterogeneity within the CFS group, the study subjects were further subdivided into those less than 25 years old ("younger CFS"), and those greater than or equal to 25 years old ("older CFS") at the time of testing. The discriminant function analysis of the RA data indicated that the 4 groups were significantly (p<0.011) different by their lipid profiles (Table 3b). The primary discriminant variables were cholesterol, which was lowest in the younger CFS, and trans-9-octadecenoic acid, which was lowest in the older CFS patients. Analysis of the PA data gave a similar result, and the canonical plot of the study subjects by the PA discriminant function (Fig. 1) shows the effective separation of the younger and older CFS patients from the control subjects. The younger and older controls could not be separated, whereas the two CFS subgroups were well resolved, indicating that the CFS group was heterogeneous in terms of lipid homeostasis. The canonical plot demonstrates the capacity of the discriminant function model to separate control subjects from CFS patients, and to further resolve the younger CFS patients from the older CFS patients. The lipid profiles were no different between males and females in either the control or CFS groups.


The percentage composition of most of the lipid saponification products from plasma samples was mainly conserved between CFS and control groups. However, multivariate analyses indicated that the CFS group had significantly different plasma lipid profiles when compared with the control group (Table 3a). These changes were primarily associated with a reduction of trans-9-octadecenoic acid (elaidic acid) and an increase in octadecenoic acid (stearic acid) in the CFS patients. The corresponding trans-9-octadecenoic acid:octadecenoic acid ratio was also found to be reduced in the CFS patients when compared with control subjects (p < 0.00002) and may represent an important marker for the objective assessment of CFS. There was no detectable trans-9-octadecenoic acid in the plasma of 25% of CFS patients compared with only 3% of control subjects (p < 0.003). These results indicated that significant changes in the metabolism of trans-9-octadecenoic acid and octadecenoic acid occured in the CFS patients.

The onset of CFS can occur at any age, but is most frequently observed among the 12-20-year-old or the 30-40-year-old age groups. The canonical plot of the lipid profile data in Fig. 1 shows that the controls can be separated from the CFS patients who comply with defined CFS ( 1), and that these CFS patients can be further subdivided into younger ( 24 years) subsets. Age was not associated with changes in the lipid profile in the controls, but was associated with changes in the PA lipid profiles in the CFS patients. The differential age-dependent changes may occur as a result of CFS or may actually represent predisposing factors for contracting and sustaining CFS. Either way, the data suggest a functional alteration of lipid metabolism/homeostasis in the CFS patients.

The separation, by discriminant function lipid profile models, of the controls from the younger and older subsets of CFS supports an organic basis to CFS and also indicates that compliance with the clinical definition of CFS [ 1] does not necessarily produce a homogeneous group of patients. Homogeneity in clinical groups is essential for the investigation of any specified illness and the development of effective treatment strategies. The use of plasma lipid profiles offers a potential means for objective assessment and characterization of CFS patients, and further investigation of CFS heterogeneity is therefore warranted.

The level of trans-9-octadecenoic acid was a major discriminant variable separating the controls from the CFS patients, as well as the younger from the older CFS subgroups.

The tissue distribution of trans-9-octadecenoic acid in humans ranges from a low of 0.3% in brain tissue to 4.0% in adipose tissue [ 19]. The daily consumption of trans-9-octadecenoic acid in the American diet is of the order of 5-7% of fatty acids consumed (reviewed in [ 20]). Since alterations in serum and tissue levels of trans-9-octadecenoic acid are profoundly influenced by alterations in diet [ 21-23], the present findings could easily be dismissed on this basis. However, the membrane concentration of trans-9-octadecenoic acid has been shown to be highly conserved in normal situations in animal models [ 21], suggesting that trans-9-octadecenoic acid may be highly regulated in certain tissues. Although increasing dietary trans-9-octadecenoic acid results in increased incorporation of the lipid into rat liver and heart [ 22], its accumulation does not normally exceed 15% of the acyl chains in any of the lipid pools, even when dietary trans-9-octadecenoic acid accounts for 46% of the fatty acids supplied in the diet. The tissue accumulation of trans-9-octadecenoic acid in rats was at the expense of octadecenoic acid and was not associated with a change in incorporation of cis-5,8,11, 14-eicosatetraenoic acid [ 23], suggesting that trans-9-octadecenoic acid behaves more like a saturated than an unsaturated fatty acid. In phospholipids, trans-9-octadecenoic acid displaces octadecenoic acid and/or hexadecanoic acid from the R1 position of phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol [ 24]. The data from this study suggest that there may have been a higher level of octadecenoic acid and a corresponding reduction of trans-9-octadecenoic acid in the R1 position of these phospholipids. Thus, while simple dietary change could be postulated as the cause of lipid changes in the present study, experimental evidence suggests that a non-dietary change in trans-9-octadecenoic acid metabolism may be occurring in CFS patients.

Increases in trans-9-octadecenoic acid have been found to inhibit cell growth and lead to increased cellular phospholipid levels [ 25]. In humans increasing amounts of trans-9-octadecenoic acid have been reported to be associated with an impairment of essential fatty acid metabolism, particularly in the C20-C22 range [ 26]. Dietary intakes of trans-9-octadecanoic acid result in a reduction of cis-5,8,11,14-eicosatetraenoic acid and an increase in cis-9,12-octadecadienoic acid in liver phosphatidylcholine and phosphatidylethanolamine [ 23]. Eighteen enzymes or cellular functions could be found that may be altered by changes in trans-9-octadecenoic acid (listed in Table 4) [ 25, 27-40]. It should be noted, however, that many of these changes were assessed in vitro, and their physiological implications have yet to be fully investigated. Several of these may be of importance in CFS as changes have been reported in carnitine acyltransferase- [ 40] and monoamine oxidase-associated metabolites [ 2].

Apart from diet, two other major possibilities exist which may offer an explanation for the reduction in trans-9-octadecenoic acid levels in CFS patients. Enzymes such as fatty acid omega-hydroxylase (cytochrome P-450ka2-CYP4A7), which may be activated to degrade prostaglandins (for example following an inflammation response), will also actively metabolize trans-9-octadecenoic acid [ 41] with a resultant reduction in trans-9-octadecenoic acid levels. It has also been reported that trans-9-octadecenoic acid has anti-viral activity [ 42] and that viruses can alter host fatty acid composition in cells [ 8]. Hence, specific viral infection might reduce trans-9-octadecenoic acid levels, thus removing a potential replication inhibitor to facilitate future reactivation and/or infection. These hypotheses may offer a basis for the increased susceptibility of certain patients to develop CFS after an infectious event, although other predisposing and environmental factors may also be implicated.


This study found that CFS patients have different plasma saponified lipid fraction profiles compared with control subjects, indicating a substantial change in lipid metabolism and homeostasis in CFS patients. These changes were primarily characterized by a reduction in trans-9-octadecenoic acid and a reduction in the trans-9-octadecenoic acid:octadecenoic acid ratio in the CFS patients. The CFS patients could be differentiated by their lipid profiles into age-based subgroups indicating lipid heterogeneity within the CFS patients.


We would like to thank Environmental Health Technologies Ltd for a grant in support of this project, Dr Peter Dobie for assisting in the clinical diagnoses, Dr John Whiting (FRACP) and Dr Ian Buttfield (FRACP, MD) for their constructive criticisms of the manuscript, and the University of Newcastle Health Service for assistance in collection of blood samples. We also thank the Gideon Lang Foundation, the Judith Mason CFS Research Grant and the Lower Hunter CFS/ME Fundraising Association for their financial support, and the members of the ME/CFS Society of NSW, particularly the Maitland support group for their support and encouragement.

(*) Dedicated to the memory of Alison Hunter

TABLE 1. Comparison of fatty acid concentrations and relative percentage abundances measured in this study compared with literature values
Legend for Chart:

C - Fatty acid
D - Control plasma mmol L-1
E - Sera et al. 1994 [18] serum mmol L-1
F - Control plasma % FA
G - Sera et al. 1994 [18] serum mmol L-1




C12:0 0.04 < 0.05 0.5% < 0.7%


C14:0 0.13 0.08-0.39 1.7% 1.1-3.6%


C16:0 2.20 1.21-3.2 29.4% 19.6-27.5%


C16:1 n-7 0.23 0.09-0.48 3.1% 1.4-3.6%


C18:0 0.62 0.45-1.07 8.3% 7.0-11.2%


C18:1 n-9 0.72 0.88-2.96 9.6% 15.7-27.2%


C18:2 n-6 1.45 1.5-3.4 19.4% 23.0-37.8%


C20:3 n-6 0.77 0.06-0.24 10.3% 1.0-3.4%


C20:4 n-6 1.33 0.31-0.84 17.8% 5.0-11.0%
TABLE 2. The relative abundance and absolute amount of the saponified lipid products detected by GC-MS in plasma extracts from control subjects (n = 39) compared with the CFS patient (n = 60) groups
Legend for Chart:

A - Saponification product
B - Common name
C - Control RA%
E - t-test p-value
F - Control mmol L-1
G - CFS mmol L-1
H - test p-value
I - Reference range in serum [18]



Polar lipid fractions

phosphoric acid

phosphate 2.78 3.17 NS


glycerol 0.76 1.05 NS

succinic acid

succinate 0.04 0.05 NS

glyceric acid

glycerate 0.18 0.21 NS


2-GP 2.59 2.76 NS


1-GP 2.00 2.17 NS

Saturated fatty acids

Dodecanoic acid (C12:0)

Lauric acid 0.07 0.08 NS 0.04
0.04 NS < 0.05

Tetradecanoic acid (C 14:0)

Myristic acid 0.62 0.69 NS 0.13
0.11 NS 0.08-0.39

Hexadecanoic acid (C16:0)

Palmitic acid 17.9 17.3 NS 2.20
1.84 NS 1.21-3.20

4-Methyl-hexadecanoic acid (4M-C 16:0)

0.08 0.08 NS 0.04
0.04 NS

Heptadecanoic acid (C17:0)

0.21 0.22 NS 0.06
0.05 NS

Octadecanoic acid (C 18:0)

Stearic acid 4.04 4.54 NS 0.62
0.55 NS 0.45-1.07

Eicosanoic acid (C20:0)

Arachidic acid 0.07 0.08 NS 0.04
0.04 NS

Docosanoic acid (C22:0)

Behenic acid 0.13 0.14 NS 0.05
0.05 NS

Tricosanoic acid (C23:0)

0.06 0.06 NS 0.04
0.04 NS

Tetracosanoic acid (C24:0)

Lignoceric acid 0.11 0.11 NS 0.05
0.04 NS

Unsaturated fatty acids

n-7 cis-7-Tetradecanoic (cis-7-C14:1)

Myristoleic acid 0.05 0.05 NS 0.04
0.04 NS

cis-9-Hexadecanoic (cis-9-C16:1)

Palmitoleic acid 1.45 1.52 NS 0.23
0.19 NS 0.09-0.48

n-9 cis-9-Octadecanoic (cis-9-C18:1)

Oleic acid 11.8 11.2 NS 0.72
0.66 NS 0.88-2.96

trans-9-Octadecanoic (trans-9-C18:1)

Elaidic acid 0.99 0.74 < 0.007 0.33
0.33 < 0.003

cis-15-Tetracosenoic (cis-15-C24:1)

Nervonic acid 0.11 0.14 NS 0.05
0.05 NS

n-6 cis-9, 12-Octadecadienoic (cis-9, 12-C18:2)

Linoleic acid 11.8 12.7 NS 1.45
1.40 NS 1.50-3.40

cis-11, 14-eicosadienoic (cis-11, 14-C20:2)

0.52 0.69 NS 0.15
0.14 NS

cis-8, 11, 14-eicosatrienoic (cis-8, 11,

DGLA 0.33 0.41 NS 0.77
0.67 NS 0.06-0.24

cis-5, 8, 11, 14-eicosatetraenoic (cis-5, 8,
11, 14-C20:4)

Arachidonic acid 0.56 0.66 NS 1.33
1.08 NS 0.31-0.84

n-3 cis-11, 14, 17-eicosatrienoic (cis-11,
14, 17-C20:3)

0.14 0.17 NS 0.27
0.31 NS

Unassigned fatty acids


1.12 1.04 NS 2.19


1.01 0.89 NS 1.90


0.15 0.18 NS 0.05


0.29 0.38 NS 0.57


0.19 0.23 NS 0.37


Cholest-5-en-3 Beta-ol

Cholesterol 37.7 36.2 NS 4.13
3.51 < 0.04 3.0-5.5[[*]]

5 Alpha-Cholest-7-en-3 Beta-ol

Lathosterol 0.06 0.05 NS 0.04
0.04 NS

Campest-5-en-3 Beta-ol

Campesterol 0.03 0.03 NS 0.04
0.04 NS

Sitost-5-en-3 Beta-ol

Sitosterol 0.02 0.02 NS 0.04
0.04 NS

Statistical Method = Hotellings T2 test and t-test; RA
Hotelling's T2 test; T2 = 7,115,000,000;
F (34,64) = 138,100,000; p < 0.00001

[*] Pathology service reference rang.
TABLE 3. Discriminant function analysis of the plasma lipid profile data (relative abundance data)
Legend for Chart:

A - Standard discriminant function model
B - Model statistics
C - Discriminant metabolites (lipid saponification products)
D - Discriminant metabolites (lipid saponification products)
E - Sensitivity/Specificity (= percentage of study subjects
correctly classified by model)
F - Sensitivity/Specificity (= percentage of study subjects
correctly classified by model)
G - Sensitivity/Specificity (= percentage of study subjects
correctly classified by model)
H - Sensitivity/Specificity (= percentage of study subjects
correctly classified by model)
I - Sensitivity/Specificity (= percentage of study subjects
correctly classified by model)
J - Sensitivity of validation group





(a) Clinical group comparisons


Control Total Total
(specificity) (accuracy) n = 39
n=39 n=99

CFS (n = 60)
Controls (n = 39)

Wilks' Lambda = 0.488 1)trans-9-octadecanoic acid
F = 1.89, p < 0.014 2)octadecanoic acid

control > CFS[a]
control < CFS 90%

82% 87% 67%

(b) Age dependent clinical subgroup comparisons

years years
n=10 n=50

24 Total
years years n = 99
n=9 n=30

a)CFS b)CFS >24 years old (n =50)
c)Controls d)Controls > 24 years old (n = 30)

Wilks' Lambda=0.160 1)Cholesterol
F= 1.48, p<0.011 2)trans-9-octadecanoic acid

a b

56% 70% 81%

[a] Indicates significant differences between mean values by
Duncan's multiple range test, p < 0.05
TABLE 4. Enzymes and metabolic functions that are altered by trans-9-octadecenoic acid
Legend for Chart:

A - Enzyme or function
B - References


Increased activity or function with
increasing trans-9-octadecenoic acid

Plasma LDL (dependent upon the cis-9, [28]
12-C18:2:C16:0 ratio)

Plasma cholesteryl ester transfer [27]
protein (CETP)

Glycerophosphate acyltransferase [25]

Delta-9 desaturase (Stearoyl-CoA desaturase) [29], [30]

Basal activity of RBC Ca2+ ATPase [31]

Decreased activity or function with
increasing trans-9-octadecenoic acid

Plasma HDL (dependent upon the cis-9, [28]
12-C18:2:C16:0 ratio)

Cyclic AMP-dependent protein kinase [32]

Steroid 5 alpha-reductase [33]

Carnitine acyltransferase [34]

Delta-5 desaturase [35]

Delta-6 desaturase [29], [30]

Diacylglycerol acyltransferase [36]

Monoamine oxidase [37]

Chymase [38]

Cardiac membrane Na+ K+ [39]

Cardiac membrane Adenylate cyclase [39]

Cardiac Beta-Adrenergic receptor numbers [39]

Thyroid hormone stimulation of RBC [31]
Ca2+ ATPase
GRAPH: FIG. 1. The canonical plot showing the separation of the older and younger CFS and control subjects achieved by the standard discriminant function model of the saponified plasma lipids (Table 3b).

[1] Fukuda K, Straus SE, Hickie I, et al. The chronic fatigue syndrome: A comprehensive approach to its definition and study. Ann Intern Med 1994; 121: 953-9.

[2] Demitrack MA, Gold PW, Dale JK, Krahn DD, Kling MA, Straus SE. Plasma and cerebrospinal fluid monoamine metabolism in patients with chronic fatigue syndrome: preliminary findings. Biol Psychiatry 1992; 32: 1065-77.

[3] Cleare AJ, Beam J, Allain T, et al. Contrasting neuroendocrine responses in depression and chronic fatigue syndrome. J Affect Disord 1995; 35: 283-89.

[4] McGregor NR, Dunstan RH, Zerbes M, Butt HL, Roberts TK, Klineberg IJ. Preliminary determination of a molecular basis to chronic fatigue syndrome. Biochem Mol Med 1996; 57: 73-80.

[5] Suhadolnik Rd, Reichenbach NL, Hitzges P, et al. Upregulation of the 2-5A Synthetase/RNase L pathway associated with chronic fatigue syndrome. Clin Infect Dis 1994; 18: S96-S104.

[6] Stayer DR, Carter WA, Brodsky I, et al. A controlled clinical trial with a specifically configured RNA drug, Poly(1)-Poly(C12U), in chronic fatigue syndrome. Clin Infect Dis 1994; 18: S88-S95.

[7] Levy JA. Viral studies of chronic fatigue syndrome. Clin Infect Dis 1994; 18: S117-S120.

[8] Behan PO, Behan WM, Horrobin D. Effect of high doses of essential fatty acids on the postviral fatigue syndrome. Acta Neurol Scand 1990; 82: 209-16.

[9] Holmes GP, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: A working case definition. Ann Intern Med 1988; 108: 387-9.

[10] Selley ML, Bartlett MR, McGuiness JA, Hapel AJ, Ardlie NG, Lacey MJ. Determination of the lipid peroxidation products trans-4-hydroxy-2-nonenal in biological samples by High Performance Liquid Chromatography and combined capillary column Gas Chromatography-Negative Ion Chemical Ionisation-Mass Spectrometry. J Chromatogr Biomed App 1989; 488: 329-40.

[11] Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 1959; 37: 911-7.

[12] Dunstan RH, Smillie RH, Grant BR. The effects of sub-toxic levels of phosphonate on the metabolism and potential virulence factors of Phytophthora palmivora. Physiol Mol Plant Path 1990; 36: 205-20.

[13] Sakoff JA, Dunstan RH, Murdoch RN. Uterine lipid alterations during early pseudopregnancy and following the artificail induction of decidualisation by concanavalin A in QS mice. Mol Reprod Dev 1996; 44: 93-102.

[14] Greenaway W, Scaysbrook T, Whatley FR. The analysis of bud exudate of Populus X euramericana and propolis, by gas chromatography mass spectrometry. Proc R Soc Lond 1987; B232: 249-72.

[15] Gurr MI, James AT. Lipid Biochemistry, an Introduction. 1971, pp. 6, 147. Chapman and Hall, London.

[16] Fogerty AC. Chemical Reactions of Lipids. In: Biochemistry and Methodology of Lipids. A.R. Johnson and J.B. Davenport (eds.), 1971, pp 29-46. Wiley-Interscience, Canada.

[17] Blaton VH, Peeters H. Integrated Approach to Plasma Lipid and Lipoprotein Analysis. In: Blood Lipids and Lipoproteins, Quantitation, Composition and Metabolism. G.J. Nelson (ed.), 1972, pp 275-314. Wiley-Interscience, Canada.

[18] Sera RK, McBride JH, Higgins, SA Rodgerson DO. Evaluation of reference ranges for fatty acids in serum. J Clin Lab Anal 1994; 8: 81-5.

[19] Adlof RO, Emken EA. Distribution of hexadecenoic, octadecenoic and octadecadienoic acid isomers in human tissue lipids. Lipids 1986; 21: 543-7.

[20] Mensink RP, Katan MB. Trans monounsaturated fatty acids in nutrition and their impact on serum lipoprotein levels in man. Prog Lipid Res 1993; 32:111-22.

[21] Masuzawa Y, Prasad MR, Lands WE. Distribution of dietary trans-octadecenoate among acyl-CoA and other lipid fractions of rat liver and heart. Biochim Biophys Acta 1987; 919: 297-306.

[22] Lanser AC, Emken EA, Ohlrogge JB. Oxidation of oleic and elaidic acids in rat and human heart homogenates. Biochim Biophys Acta 1986; 875: 510-5.

[23] Lawson LD, Hill EG, Holman RT. Dietary fats containing concentrates of cis or trans octadecenoates and the patterns of polyunsaturated fatty acids of liver phosphatidylcholine and phosphatidylethanolamine. Lipids 1985; 20: 262-7.

[24] Wolff RL, Entressangles B. Steady-state fluorescence polarization study of structurally defined phospholipids from liver mitochondria of rats fed elaidic acid. Biochim Biophys Acta 1994; 1211: 198-206.

[25] Urade R, Kito M. Phospholipid metabolism in V79-R membranes composed of phospholipid molecular species containing trans-monoenoic fatty acids. J Biochem (Tokyo) 1984; 96: 821-8.

[26] Koletzko B. Trans fatty acids may impair biosynthesis of long-chain polyunsaturates and growth in man. Acta Paediatr 1992; 81: 302-6.

[27] Abbey M, Nestel PJ. Plasma cholesteryl ester transfer protein activity is increased when trans-elaidic acid is substituted for cis-oleic acid in the diet. Atherosclerosis 1994; 106: 99-107.

[28] Nestel PJ, Noakes M, Belling GB, McArthur R, Clifton RM, Abbey M. Plasma cholesterol-lowering potential of edible-oil blends suitable for commerical use. Am J Clin Nutr 1992; 55: 46-50.

[29] De Schrijver R, Privett OS. Interrelationship between dietary trans fatty acids and the 6- and 9-desaturases in the rat. Lipids 1982; 17: 27-34.

[30] Rosenthal MD, Whitehurst MC. Selective effects of isomeric cis and trans fatty acids on fatty acyl delta 9 and delta 6 desaturation by human skin fibroblasts. Biochim Biophys Acta 1983; 753: 450-9.

[31] Davis FB, Davis PJ Bias SD, Schoenl M. Action of long-chain fatty acids in vitro on Ca2 + -stimulatable, Mg2 +- dependent ATPase activity in human red cell membranes. Biochem J 1987; 248: 511-6.

[32] Doolan CM, Keenan AK. Inhibition by fatty acids of cyclic AMP-dependent protein kinase activity in brush border membranes isolated from human placental vesicles. Br J Pharmacol 1994; 111: 509-14.

[33] Liang T, Liao S. Inhibition of steroid 5 alpha-reductase by specific aliphatic unsaturated fatty acids. Biochem J 1992; 285: 557.-62.

[34] Ide T, Watanabe M, Sugano M, Yamamoto I. Activities of liver mitochondrial and peroxisomal fatty acid oxidation enzymes in rats fed trans fat. Lipids 1987; 22: 6-10.

[35] Rosenthal MD, Doloresco MA. The effects of trans fatty acids on fatty acyl delta 5 desaturation by human skin fibroblasts. Lipids 1984; 19: 869-74.

[36] Hillmar I, Henze K, Zollner N. Influence of fatty acids on microsomal diacylglycerol acyltransferase activity in primary cultures of hepatocytes. Hoppe Seylers Z Physiol Chem 1983; 364: 173-8.

[37] Kritchevsky D. Trans fatty acid effects in experimental atherosclerosis. Fed Proc 1982; 41: 2813-7.

[38] Kido H, Fukusen N, Katunuma N. Inhibition of chymase activity by long chain fatty acids. Arch Biochem Biophys 1984; 230: 610-4.

[39] Alam SQ, Ren YF, Alam BS. Effect of dietary trans fatty acids on some membrane-associated enzymes and receptors in rat heart. Lipids 1989; 24: 39--44.

[40] Kuratsune H, Yamaguti K, Takahashi M, Misaki H, Tagawa S, Kitani T. Acylcarnitine deficiency in chronic fatigue syndrome. Clin Infect Dis 1994; 18: S62-S67.

[41] Sawamura A, Kusunose E, Satouchi K, Kusunose M. Catalytic properties of rabbit kidney fatty acid omega-hydroxylase cytochrome P-450ka2 (CYP4A7). Biochim Biophys Acta 1993; 1168: 30-6.

[42.] Horowitz B, Piet MP, Prince AM, Edwards CA, Lippin A, Walakovits LA. Inactivation of lipid-enveloped viruses in labile blood derivatives by unsaturated fatty acids. Vox Sang 1988; 54: 14-20.


By R. Hugh Dunstan, DPhil, Neurobiology Research Unit, Centre for Oral Health Research, University of Sydney, Westmead Hospital, Westmead, NSW 2084, Australia; Neil R. McGregor, MDSc, Neurobiology Research Unit, Centre for Oral Health Research, University of Sydney, Westmead Hospital, Westmead, NSW 2084, Australia; Jennifer A. Watkins, BSc, Environmental Medical Centre, Mosman, Sydney, NSW 2088, Australia; Mark Donohoe, MB BSc, Environmental Medical Centre, Mosman, Sydney, NSW 2088, Australia; Raymond N. Murdoch, PhD, Environmental Medical Centre, Mosman, Sydney, NSW 2088, Australia; Warren G. Taylor, BSc, Environmental Medical Centre, Mosman, Sydney, NSW 2088, Australia; Timothy K. Roberts, PhD, and Henry L. Butt, PhD, Division of Microbiology and Infectious Diseases, Hunter Area Pathology Service, John Hunter Hospital, Newcastle, NSW 2305, Australia; Raymond N. Murdoch, PhD, Bioanalytical Research Group, Department of Biological Sciences, The University of Newcastle, Callaghan, NSW 2308, Australia and Warren G. Taylor, BSc, Bioanalytical Research Group, Department of Biological Sciences, The University of Newcastle, Callaghan, NSW 2308, Australia

Adapted by DPhil , MDSc , BSc , MB BSc , PhD , BSc , PhD , PhD , PhD and BSc

Share this with your friends