Evidence from in vitro and animal studies suggests that conjugated linoleic acid (CLA), a group of positional and geometric conjugated dienoic isomers of linoleic acid, modulates immune function (for a review, see O'Shea et al. (Reference O'Shea, Bassaganya-Riera and Mohede1)). CLA has been shown to affect both humoral and cellular immune responses. Lymphocyte proliferation, immunoglobulin and cytokine production as well as proportion of T cell subtypes have been affected by CLA supplementation in animals(Reference O'Shea, Bassaganya-Riera and Mohede1).
CLA may also modulate the balance between Th1- and Th2-type immune responses by decreasing production of Th2 cytokines (IL-4, IL-5)(Reference Yang and Cook2–Reference Yamasaki, Kitagawa and Chujo4) and by increasing production of Th1 cytokines (IL-2, interferon (IFN)-γ)(Reference Yamasaki, Kitagawa and Chujo4–Reference Hayek, Han, Wu, Watkins, Meydani, Dorsey, Smith and Meydani7). In animals, CLA has shown inhibitory effects in type 1 hypersensitivity reactions(Reference Ishiguro, Oku, Suitani and Yamamoto8, Reference Whigham, Cook, Stahl, Saban, Bjorling, Pariza and Cook9). CLA suppressed antigen-induced anaphylactic reactions in mice by inhibiting the decrease in blood pressure and blood flow(Reference Ishiguro, Oku, Suitani and Yamamoto8). Both orally and topically given CLA also suppressed scratching behaviour and vasodilation induced by a histamine release agent, compound 48/80(Reference Ishiguro, Oku, Suitani and Yamamoto8). In a guinea-pig tracheal superfusion model, CLA supplementation significantly reduced antigen-induced histamine and PGE2 release(Reference Whigham, Cook, Stahl, Saban, Bjorling, Pariza and Cook9).
In human subjects, effects of CLA have been modest. Antibodies to hepatitis B vaccination were increased in subjects supplemented daily with 1·7 g cis-9, trans-11-CLA and trans-10, cis-12-CLA (1:1), but no changes in cytokine production, lymphocyte proliferation, natural killer cell activity or delayed hypersensitivity were seen(Reference Albers, van der Wielen, Brink, Hendriks, Dorovska-Taran and Mohede10). Effects of CLA did not differ from those of linoleic acid on lymphocyte proliferation and cytokine production in another study(Reference Nugent, Roche, Noone, Long, Kelleher and Gibney11). However, significant changes in immunoglobulin and cytokine production were observed in subjects fed 3 g CLA mixture (1:1 cis-9, trans-11-CLA and trans-10, cis-12-CLA) daily(Reference Song, Grant, Rotondo, Mohede, Sattar, Heys and Wahle12). IgA and IgM production increased, while IgE, TNF-α and IFN-γ decreased(Reference Song, Grant, Rotondo, Mohede, Sattar, Heys and Wahle12). Purified cis-9, trans-11- and trans-10, cis-12-CLA isomers decreased mitogen-induced T lymphocyte activation dose-dependently, but did not affect ex vivo cytokine production(Reference Tricon, Burdge, Kew, Banerjee, Russell, Grimble, Williams, Calder and Yaqoob13).
All human studies on the effects of CLA on immune function have thus far been conducted in healthy subjects. The effects of CLA could be more profound in situations of immune imbalance, such as allergy. Thus, we conducted a study on the immunological and metabolic effects of cis-9, trans-11-CLA in subjects with birch pollen allergy. A CLA mixture containing predominantly the cis-9, trans-11 isomer was chosen since, should it prove beneficial, its dietary intake from dairy products and beef can be modulated by increasing the CLA content of these products through altered cow feeding regimens(Reference Lock and Bauman14). Also, adverse effects have been reported for the other major isomer, trans-10, cis-12-CLA(Reference Riserus, Arner, Brismar and Vessby15–Reference Riserus, Vessby, Arner and Zethelius17).
Subjects and methods
Forty subjects (twenty-eight women and twelve men; age 20–46 years) with a diagnosis of birch pollen allergy were recruited from students and employees of the University of Helsinki. Health status was determined in screening tests. Exclusion criteria included chronic medication, consumption of dietary supplements, heavy smoking (more than ten cigarettes/d), heavy alcohol consumption (more than ten units per week), obesity (BMI>30 kg/m2), pregnancy, lactation, diagnosis of CVD, diabetes, liver or endocrine dysfunction, asthma or other chronic inflammatory disease besides allergy. The selected subjects were normotensive (blood pressure < 140/90 mmHg), normolipidaemic (total cholesterol < 6 mmol/l, TAG concentration < 2·0 mmol/l), normoglycaemic (blood glucose < 6·0 mmol/l) and had no signs of acute inflammation at baseline (C-reactive protein < 10 mg/l).
A randomised, placebo-controlled study design was used. Subjects were assigned into CLA and placebo groups (twenty per group) and consumed three 1000 mg capsules daily for 12 weeks, providing 2 g cis-9, trans-11-CLA or high-oleic acid sunflower-seed oil. The CLA capsules contained 65·3 % cis-9, trans-11-CLA, 13·8 % 18 : 1n-9, 8·5 % trans-10, cis-12-CLA, 1·8 % cis-9, cis-12-18 : 2, 0·7 % trans-, trans-CLA and 5·4 % other fatty acids. The placebo capsules contained 78·1 % 18 : 1n-9, 12·7 % cis-9, cis-12-18 : 2, 0·1 % 18 : 3 and 8·6 % other fatty acids. Capsules were provided by Loders Croklaan, Lipid Nutrition (Wormerveer, The Netherlands).
Subjects were free living and were asked to maintain their normal dietary habits and physical activity. Allergy medication (antihistamine, eye drops, nasal spray) was provided during the birch pollen season, to be used when necessary for allergy symptoms. Subjects recorded allergy symptoms on a scale from 1 to 3 (1 = mild symptoms, 3 = severe symptoms) and use of allergy medication (as number of tablets, eye drops or sprays) in their study diaries daily.
Fasting blood samples were taken at baseline (week 0) and at the end of the pollen season (week 12). Morning urine samples were collected at the same time points.
Laboratory analyses
Blood count, lipids, glucose, C-reactive protein
Total blood count was performed on samples from week 0 and 12 using a Micros 60 blood analyser (ABX Diagnostics, Montpellier, France). Plasma lipid, glucose and C-reactive protein analyses were performed on a Konelab 20 autoanalyser (Thermo Fischer, Vantaa, Finland).
A commercial enzyme immunassay kit (Biosource, Nivelles, Belgium) was used to analyse plasma insulin concentrations. Samples were analysed in duplicate, with all samples of one subject on the sample plate. Control samples included in the kit were analysed with each batch of samples.
The homeostasis model for insulin resistance was calculated from fasting insulin and glucose concentrations ((insulin0 × glucose0)/22·5)(Reference Matthews, Hosker, Rudenski, Naylor, Treacher and Turner18) and insulin sensitivity was calculated using the revised quantitative insulin sensitivity check index (1/(log glucose0+log insulin0+log NEFA0))(Reference Perseghin, Caumo, Caloni, Testolin and Luzi19).
Lymphocyte subsets
Proportions of CD4+ and CD8+ cells were determined from whole blood (100 μl) with fluorescent labelled mouse monoclonal antibodies against CD4 and CD8 (Serotec, Oxford, Oxon, UK). Samples were first incubated with the fluorescein isothiocyanate-conjugated antibody (10 μl) for 15 min at room temperature after which the erythrocytes were lysed with Cellkit C-04 (Cellset, Galmiz, Switzerland) and the cells were washed twice with PBS. Lymphocytes were fixed and the percentage of CD4+ and CD8+ cells was determined from a total of 104 cells with flow cytometry using a FACScan (Becton Dickinson, Franklin Lakes, NJ, USA).
Plasma immunoglobulins
Plasma concentrations of IgA, IgG and IgM were determined using an automated blood analyser (Konelab, Espoo, Finland). Total and birch-specific IgE concentrations were analysed using the Pharmacia CAP FEIA method (Pharmacia Diagnostics AB, Uppsala, Sweden).
Peripheral blood mononuclear cells stimulation
Peripheral blood mononuclear cells (PBMC) were isolated from 9 ml heparin blood by Ficoll-Paque centrifugation(Reference Miroli, James and Spitz20). Blood was diluted with 10 % fetal calf serum–Roswell Park Memorial Institute (FCS/RPMI) culture media (1:1). Cells were counted in a Neubauer Improved counting chamber and diluted in the culture media to achieve a concentration of 2 × 106/ml. Cells were then frozen at − 70°C.
PBMC were stimulated using birch pollen extract (ALK, Hørsholm, Denmark). Pollen extract was diluted using 10 % FCS/RPMI media to a final concentration of 10 000 standard quality (SQ)/ml.
Diluted antigen was pipetted into ninety-six-well microplates and cell suspensions (2 × 106/ml) were added to each well containing antigen or blank (10 % FCS/RPMI). Each sample was analysed in triplicate. The plates were incubated at 37°C and 5 % CO2 for 6 d. Samples were centrifuged and supernatant fractions were collected. Supernatant fractions of replicate wells were pooled and stored at − 70°C until analysed.
Cytokine production
In vitro production of IL-6 and TNF-α by stimulated PBMC were determined as general markers of inflammation. Plasma granulocyte macrophage colony-stimulating factor (GM-CSF), eosinophil-derived neurotoxin (EDN) and eosinophil cationic protein (ECP) concentrations as well as in vitro production of IL-4, IL-5, IL-10, IL-13 and IFN-γ by stimulated PBMC were analysed as markers of allergic inflammation. Plasma concentrations of EDN and ECP were determined using kits from MBL (Nagoya, Japan) according to the manufacturer's instructions. All other cytokines were analysed using ELISA kits from R&D Systems (Minneapolis, MN, USA). All samples were analysed in duplicate. Control samples included in each kit were measured with each batch of samples. The inter-assay and intra-assay CV were ≤ 10 % for all cytokines, excluding the inter-assay CV for IL-13 (11 %) and GM-CSF (15 %).
Lipid peroxidation
Urine samples collected before and after supplementation were analysed for free 8-iso-PGF2α, an indicator of free radical-mediated lipid peroxidation and 15-keto-dihydro-PGF2α, an indicator of cyclo-oxygenase-catalysed lipid peroxidation. Unextracted urine was analysed using highly specific and sensitive RIA for free 8-iso-PGF2α and 15-keto-dihydro-PGF2α as previously described by Basu(Reference Basu21, Reference Basu22). Urinary concentrations of 8-iso-PGF2α and 15-keto-dihydro-PGF2α were adjusted for creatinine values measured with a commercial kit (IL test; Monarch Instrument, Amherst, NH, USA).
Statistical analysis
Data that were not normally distributed were log-transformed before testing. Differences between the two treatment groups were analysed with the unpaired t test. Spearman correlation tests were used to investigate relationships between the cytokines. Values of P < 0·05 were considered statistically significant.
Results
Compliance
Thirty-eight subjects completed the study. One subject withdrew from the CLA group due to health problems considered unrelated to the supplement and one subject from the placebo group due to personal reasons.
Compliance according to capsule counting was 98 % in both groups. Also, analysis of plasma fatty acids indicated good compliance. In the CLA groups, the concentration of cis-9, trans-11-CLA increased from 0·20 % to 0·75 % of total fatty acids in plasma (P < 0·001), while no change was seen in the placebo group (0·19 v. 0·22 %; P = 0·660) (data not shown).
Allergy symptoms and medication
The CLA group reported a better overall feeling of wellbeing (P < 0·05) (data not shown). There were no significant differences between groups in eye symptoms or total allergy symptoms (Fig. 1). Subjects in the CLA group reported significantly less sneezing (P < 0·05) and tendency to less nasal symptoms (P = 0·07). Use of medication did not differ between groups (data not shown) and correlated with allergy symptoms in both groups (r 0·698, P = 0·01 for the CLA group and r 0·521, P = 0·02 for the control group).
Fig. 1 Allergy symptoms in subjects supplemented with cis-9, trans-11-conjugated linoleic acid (□) or placebo (■). Values are means, with their standard deviations represented by vertical bars.
Blood count, lymphocyte subsets, lipids, glucose
No significant changes occurred in blood count, lymphocyte subsets, plasma lipids, glucose, insulin, homeostasis model for insulin resistance or quantitative insulin sensitivity check indices during the intervention (data not shown).
Plasma immunoglobulins
CLA supplementation significantly decreased plasma IgA concentrations (P < 0·01) (Table 1). No changes in plasma IgG or IgM were observed. Concentrations of total and birch-specific IgE increased in both groups during the intervention. The increase in birch-specific IgE tended to be smaller in the CLA group (P = 0·07).
Table 1 Plasma immunoglobulins in subjects supplemented with cis-9, trans-11-conjugated linoleic acid (CLA) or placebo
(Mean values and standard deviations)
* P for difference between the CLA and placebo groups.
In vitro cytokine production
Changes in the in vitro production of cytokines are presented in Fig. 2. The production of IL-4, IL-5, IL-6, IL-10, IL-13 and IFN-γ increased in both the CLA and placebo groups during the study. Production of TNF-α increased in the placebo group, but remained unchanged in the CLA group (P < 0·01 for difference between groups).
Fig. 2 In vitro production of TNF-α (A), interferon-γ (B), IL-4 (C), IL-5 (D), IL-6 (E), IL-10 (F) and IL-13 (G) by stimulated peripheral blood mononuclear cells in subjects supplemented with cis-9, trans-11-conjugated linoleic acid (□) or placebo (■). Values are means, with their standard deviations represented by vertical bars. * Mean value is significantly different from that of the placebo group (P < 0·05).
IL-4 production increased 15–20-fold during the intervention, with no differences between groups.
IL-5 production was lower in the CLA group compared with placebo (P < 0·05), whereas IL-10 production tended to be higher (P = 0·07). Production of both IL-6 and IL-13 increased approximately twofold, with no differences between the CLA and placebo groups. Also IFN-γ increased in both groups, but the increase was greater in the placebo group (P < 0·05 for difference between groups).
Several significant correlations between the cytokines were observed. IL-4 production was correlated with IL-5 (r 0·635; P < 0·001) and IL-13 (r 0·470; P < 0·01). IL-5 production was also correlated with IL-10 (r 0·635; P < 0·05) and IL-13 (r 0·474; P < 0·01). IL-6 and TNF-α were highly correlated (r 0·765; P < 0·01). IFN-γ was correlated with IL-6 (r 0·459; P < 0·01) and TNF-α (r 0·611; P < 0·001).
TNF-α and IFN-γ were also correlated with sneezing and total symptoms, whereas no associations between IL-4, IL-5, IL-6, IL-10 or IL-13 and allergy symptoms were found.
Inflammatory markers and lipid peroxidation
Plasma C-reactive protein concentrations did not differ between groups (data not shown). However, a significant difference between groups in plasma GM-CSF was detected (P < 0·05), as GM-CSF remained unchanged in the CLA group, but increased significantly (P = 0·02) in the placebo group during the pollen season (Table 2). Changes in GM-CSF correlated with nasal symptoms (r 0·361; P = 0·031), changes in IL-4 (r 0·361; P = 0·032), IL-5 (r 0·440; P = 0·007) and IFN-γ (r 0·558; P < 0·001). Also plasma EDN concentrations increased significantly in the placebo group (P = 0·04), but remained unchanged in the CLA group (P = 0·046 for difference between groups) (Table 2). A similar tendency was seen in plasma ECP concentrations (P = 0·09 for difference between groups).
Table 2 Effects of supplementation with cis-9, trans-11-conjugated linoleic acid (CLA) or placebo on plasma cytokine and urinary isoprostane concentrations
(Mean values and standard deviations)
GM-CSF, granulocyte macrophage colony-stimulating factor; ECP, eosinophil cationic protein; EDN, eosinophil-derived neurotoxin.
* P for difference between the CLA and placebo groups.
On the contrary, urinary excretion of both 8-iso-PGF2α (P < 0·01) and 15-keto-dihydro-PGF2α (P < 0·05) increased significantly in the CLA group, but remained unchanged in the placebo group (Table 2). Changes in 8-iso-PGF2α (r 0·467; P = 0·003) and 15-keto-dihydro-PGF2α (r 0·381; P = 0·020) correlated with eye symptoms, whereas no correlations were found between isoprostane excretion and other allergy symptoms or cytokine production.
A summary of the changes in allergy symptoms and immune responses observed in response to CLA supplementation is presented in Table 3.
Table 3 Summary of the effects of conjugated linoleic acid (CLA) supplementation on allergy symptoms and immune responses in subjects with birch pollen allergy
–, No effect; ↓ , decreased (P < 0·05 for difference between CLA and placebo group); IFN, interferon; ↑ , increased (P < 0·05 for difference between CLA and placebo group); ECP, eosinophil cationic protein; EDN, eosinophil-derived neurotoxin; GM-CSF, granulocyte macrophage colony-stimulating factor.
Discussion
In our allergic subjects, cis-9, trans-11-CLA supplementation reduced sneezing, production of TNF-α, IFN-γ, IL-5, EDN and GM-CSF compared with high-oleic acid sunflower-seed oil. Previously, no data exist on the effects of CLA in allergic humans.
It is unclear why specifically sneezing was affected in our subjects, but it may be related to changes in cytokine production. Total allergy symptoms and sneezing correlated with the in vitro production of TNF-α and IFN-γ, both of which were reduced by CLA. Previously, scratching behaviour and vasodilation of the skin were inhibited in mice by topically applied CLA(Reference Ishiguro, Oku, Suitani and Yamamoto8). In subjects infected with rhinovirus, 2 g CLA/d supplementation (cis-9, trans-11-CLA and trans-10, cis-12-CLA; 80:20) for 4 weeks before rhinovirus inoculation reduced coughing and sore throat compared with placebo(Reference Hayden and O'Shea23). Associations with cytokine production were not reported.
The decreased concentrations of TNF-α and IFN-γ suggest that cis-9, trans-11-CLA has inhibitory effects on Th1 cytokine production. TNF-α is an acute-phase protein and a key mediary in the local immune response. Decreased TNF-α in serum and tissues of animals has been reported after CLA supplementation in several studies(Reference Yang and Cook2, Reference Changhua, Jindong, Defa, Lidan, Shiyan and Jianjun24–Reference Akahoshi, Goto and Murao26). Song et al. (Reference Song, Grant, Rotondo, Mohede, Sattar, Heys and Wahle12) also detected a 20 % decrease in TNF-α in healthy subjects supplemented with 3 g CLA mixture/d (1:1 cis-9, trans-11-CLA and trans-10, cis-12-CLA). However, Albers et al. (Reference Albers, van der Wielen, Brink, Hendriks, Dorovska-Taran and Mohede10) did not observe changes in TNF-α production following supplementation with 1·7 g CLA mixture/d in healthy subjects. Also, Tricon et al. (Reference Tricon, Burdge, Kew, Banerjee, Russell, Grimble, Williams, Calder and Yaqoob13) found no consistent effects for daily 0·6–2·5 g cis-9, trans-11-CLA or trans-10, cis-12-CLA on TNF-α production in healthy men. A possible explanation for the discrepancy between the studies is that due to the low baseline cytokine production in healthy subjects, significant changes are difficult to detect. Similarly, n-3 fatty acids have little effect in healthy subjects, whereas studies in different inflammatory conditions have shown dose-dependent anti-inflammatory effects for EPA and DHA(Reference Sijben and Calder27).
IFN-γ has diverse immunoregulatory effects. It promotes Th1 responses, while inhibiting development of Th2 clones and IL-4-induced IgE synthesis(Reference Billiau and Dijkmans28). However, recent evidence suggests that IFN-γ may also contribute to chronic inflammation(Reference Cho, Stanciu, Holgate and Johnston29, Reference Litonjua, Sparrow, Guevarra, O'Connor, Weiss and Tollerud30). In the present study, CLA supplementation decreased IFN-γ production. In line with the present results, Song et al. (Reference Song, Grant, Rotondo, Mohede, Sattar, Heys and Wahle12) also observed a 35 % decrease in IFN-γ in healthy subjects after daily supplementation with 3 g CLA mixture. In animals, CLA supplementation has increased IFN-γ production in healthy mice(Reference Yamasaki, Kitagawa and Chujo4), but decreased IFN-γ in bacterial- or viral-induced inflammation in pigs(Reference Bassaganya-Riera, Pogranichniy, Jobgen, Halbur, Yoon, O'Shea, Mohede and Hontecillas5, Reference Hontecillas, Wannemeulher, Zimmermann, Hutto, Wilson, Ahn and Bassaganya-Riera31). Both in animals and in our subjects, the decrease in IFN-γ was also associated with decreased symptoms of the disease. The physiological significance of decreased IFN-γ in allergic subjects in response to CLA supplementation remains to be resolved.
Since eosinophil activation plays a major role in allergic inflammation, several factors affecting or indicating eosinophil activation were analysed. Of these, IL-5, GM-CSF and EDN were modulated by CLA supplementation in the present study, whereas IL-13 and ECP were not affected (Table 3). CLA has decreased IL-4 concentrations in animals(Reference Yang and Cook2–Reference Yamasaki, Kitagawa and Chujo4), but effects on IL-5 and IL-13 have been reported rarely. Yamasaki et al. (Reference Yamasaki, Chujo and Hirao32) supplemented C57BL/6J mice purified cis-9, trans-11-CLA, trans-10, cis-12-CLA, or a 50:50 blend of these two isomers for 3 weeks. Supplementation with the cis-9, trans-11-CLA isomer increased concanavalin A-stimulated TNF-α production, whereas no differences were observed between the groups with respect to IL-2, IL-4, IL-5 and IFN-γ production. In another study by the same group(Reference Yamasaki, Kitagawa and Chujo4), the effects of CLA were dependent on the form of CLA: production of IL-5 and IL-13 was up regulated over 2-fold in splenocytes of mice fed 1 % CLA as TAG, but production of both was down regulated when the CLA was fed as NEFA as in the present study.
CLA supplementation inhibited the increase in EDN and GM-CSF concentrations seen in the placebo group during the pollen season. No correlations were seen between EDN and allergy symptoms, whereas changes in GM-CSF concentrations were correlated with nasal symptoms. Indications of an inhibitory effect of cis-9, trans-11-CLA on eosinophil activation were also found in a recent study by Jaudszus et al. (Reference Jaudszus, Foerster, Kroegel, Wolf and Jahreis33). Cis-9, trans-11-CLA, but not trans-10, cis-12-CLA, inhibited ECP production in a cytokine-stimulated co-culture of human eosinophils and bronchial epithelial cells(Reference Jaudszus, Foerster, Kroegel, Wolf and Jahreis33). Also, expression of eosinophil activation markers CD69 and CD13, which are induced by IL-5 and IL-13, was reduced by cis-9, trans-11-CLA(Reference Jaudszus, Foerster, Kroegel, Wolf and Jahreis33). Effects of CLA on GM-CSF or EDN have not been reported previously. Thus, more studies are needed to confirm our observations. Since local effects at the site of inflammation may be more pronounced than systemic effects, in the present study for example, analysis of nasal fluid could have given more insight into the effects of CLA on eosinophil activation and cytokine production.
In addition to total and birch-specific IgE, also IgA, IgG and IgM were measured, since previous studies both in animals and human subjects have shown that CLA modulates immunoglobulin production. CLA supplementation increased IgA, IgG and IgM production and decreased IgE production in rat spleen lymphocytes with only 0·1 % CLA in the diet(Reference Yamasaki, Chujo and Hirao32, Reference Sugano, Tsujita, Yamasaki, Noguchi and Yamada34, Reference Yamasaki, Kishihara, Mansho, Ogino, Kasai, Sugano, Tachibana and Yamada35). In pigs, 0·5 % CLA increased IgG concentrations(Reference Bontempo, Sciannimanico, Pastorelli, Rossi, Rosi and Corino36). We, however, observed a decrease in IgA and a similar tendency for IgE in the CLA group, suggesting a suppressive effect on immunoglobulin production. The present results are also in contrast with Song et al. (Reference Song, Grant, Rotondo, Mohede, Sattar, Heys and Wahle12), who observed an increase in plasma IgA and IgM in healthy subjects fed 3 g CLA/d. A decrease in IgE was observed in both studies. The discrepancy between the present results and those of others may be due to the fact that in all other studies the CLA supplement contained trans-10, cis-12-CLA in nearly equal amounts as cis-9, trans-11-CLA.
An increase in free radical- and cyclo-oxygenase-mediated lipid oxidation after CLA supplementation has been observed in several studies(Reference Basu, Smedman and Vessby37–Reference Taylor, Williams, Rhys, James and Frenneaux40). In most studies, a CLA mixture containing nearly equal proportions but different doses of the cis-9, trans-11 and trans-10, cis-12 isomers(Reference Riserus, Basu, Jovinge, Fredrikson, Ärnlöv and Vessby16, Reference Basu, Smedman and Vessby37, Reference Basu, Riserus, Turpeinen and Vessby38, Reference Taylor, Williams, Rhys, James and Frenneaux40) or pure trans-10, cis-12-CLA(Reference Riserus, Basu, Jovinge, Fredrikson, Ärnlöv and Vessby16) has been used, whereas also cis-9, trans-11-CLA (containing 7·3 % trans-10, cis-12-CLA) increased 8-iso-PGF2α excretion by 50 % in obese men(Reference Riserus, Vessby, Ärnlöv and Basu39). Previously, we detected an increase in urinary 8-iso-PGF2α after dietary supplementation with vaccenic acid (trans-11-C18 : 1), a precursor of cis-9, trans-11-CLA(Reference Turpeinen, Mutanen, Aro, Salminen, Basu, Palmquist and Griinari41). A study using pure CLA isomers reported a significantly greater isoprostane production by the trans-10, cis-12-CLA isomer than cis-9, trans-11-CLA(Reference Riserus, Basu, Jovinge, Fredrikson, Ärnlöv and Vessby16), which may reflect the different induction capacity of arachidonic acid oxidation by the two CLA isomers. Interestingly, the increase in lipid peroxidation in response to CLA (50 % for cis-9, trans-11-CLA; 578 % for trans-10, cis-12-CLA)(Reference Riserus, Basu, Jovinge, Fredrikson, Ärnlöv and Vessby16, Reference Riserus, Vessby, Ärnlöv and Basu39) is of a different magnitude from the effect of other dietary fatty acids, for example, linoleic acid (42 % increase)(Reference Turpeinen, Basu and Mutanen42) or fish oils (20–27 % increase)(Reference Mori, Puddey, Burke, Croft, Dunstan, Rivera and Beilin43), and significantly greater than that caused by smoking (about 100 % increase), a strong inducer of oxidative stress(Reference Helmersson, Larsson, Vessby and Basu44). Nevertheless, despite the induced free radical- and cyclo-oxygenase-mediated lipid oxidation, other inflammatory markers measured in the present study do not suggest a pro-inflammatory effect for cis-9, trans-11-CLA. Excretion of both 8-iso-PGF2α and 15-keto-dihydro-PGF2α correlated with eye symptoms, but this observation should be confirmed in other studies to determine whether an increase in lipid oxidation has clinical significance in allergic subjects.
The effects of CLA seen in the present study resemble the anti-inflammatory effects reported for long-chain n-3 fatty acids. EPA and DHA have decreased lymphocyte proliferation and cytokine production in a large number of studies(Reference Sijben and Calder27). The discrepancy between our observations in allergic subjects and reported effects of CLA in healthy subjects is also in line with findings with long-chain n-3 fatty acids, i.e. healthy subjects are rather insensitive to the effects of n-3 fatty acids, whereas effects are more profound in different inflammatory conditions(Reference Sijben and Calder27). Evidence regarding the immune-modulating effects of n-6 fatty acids is more controversial. Linoleic acid is generally considered pro-inflammatory as it is metabolised into arachidonic acid, which in turn is the precursor of the pro-inflammatory eicosanoids PGE2 and leucotriene B4. However, another n-6 fatty acid, γ-linolenic acid, has been associated with anti-inflammatory effects mainly in atopic dermatitis, but the evidence is not consistent(Reference Van Gool, Zeegers and Thijs45).
We observed inhibitory effects on both Th1 and Th2 responses and specifically eosinophil activation, whereas some established markers of allergy (IgE, IL-4) were not affected (Table 3). Recent studies suggest that both Th1 and Th2 responses are increased in allergic subjects(Reference Smart and Kemp46, Reference Ng, Holt and Prescott47), contrary to an earlier hypothesis of an impaired Th1 function and a Th2 skewing effect. Thus, suppression of both Th1 and Th2 cytokine responses seen in the present study suggest a general anti-inflammatory effect for cis-9, trans-11-CLA. The effects of CLA could be mediated via several mechanisms of action. Yu et al. (Reference Yu, Correll and Vanden Heuvel48) and Jaudszus et al. (Reference Jaudszus, Foerster, Kroegel, Wolf and Jahreis33) reported that the decrease in pro-inflammatory cytokines (TNF-α, IL-6, IL-8) by CLA was mediated via PPARγ activation. CLA may also modulate cytokine synthesis by directly affecting gene expression of cytokines(Reference Whigham, Higbee, Bjorling, Park, Pariza and Cook49) or via changes in eicosanoid synthesis and signalling(Reference Whigham, Higbee, Bjorling, Park, Pariza and Cook49, Reference Pariza, Park and Cook50). As the present study is the first to study effects of cis-9, trans-11-CLA on immune responses in allergic subjects, more studies in allergy and asthma are warranted to confirm our observations.
Acknowledgements
The subjects are gratefully acknowledged for their participation. A. M. T. designed the study, recruited, screened and sampled the volunteers and conducted the experimental work. N. Y., E. v W. and S. B. analysed or supervised analysis of the samples. A. A. provided consultation on medical issues. A. M. T. wrote the manuscript with input from all authors. None of the authors had any financial or personal interest in any company or organisations sponsoring the research, including advisory board affiliations.
The study was funded by the Academy of Finland and Loders Croklaan (Wormerveer, The Netherlands). The capsules used in the study were a gift from Loders Croklaan, Lipid Nutrition (Wormerveer, The Netherlands).
