Conjugated linoleic acid and cancer

Since the initial identification of CLA from grilled minced beef and its anticarcinogenic effects on skin cancer tumours in mice⁽Reference Ha, Grimm and Pariza⁷⁾, the intervening years have provided a cascade of studies and reviews examining the anticarcinogenic properties of CLA. The mechanisms relating to anticarcinogenic properties of CLA are largely unresolved; CLA may act by antioxidant mechanisms, pro-oxidant cytotoxicity, inhibition of nucleotide and protein synthesis, reduction of cell proliferative activity and inhibition of both DNA–adduct formation and carcinogen activation⁽Reference Parodi^17–Reference Kelley, Hubbard and Erickson¹⁹⁾. The studies examined in these reviews have identified potential beneficial effects of CLA on colorectal, breast and prostate cancer, with the majority of evidence from animal and in vitro studies.

CLA has shown consistent anticarcinogenic effects against several types of experimental cancer⁽Reference Belury²⁰⁾ including breast cancer⁽Reference Ip, Chin and Scimeca^21, Reference Ip, Singh and Thompson²²⁾. A review by Kelley et al. ⁽Reference Kelley, Hubbard and Erickson¹⁹⁾ examined the literature in terms of the effects of studies where purified isomers of CLA were administered. Results from in vitro studies suggest that the effects of the two isomers of CLA vary according to the cancer model examined. In the majority of studies, c9, t11-CLA did not inhibit tumour growth, whereas t10, c12-CLA demonstrated inhibitory effects in studies using mouse and human mammary tumour cell lines. The t10, c12-CLA isomer also inhibited cell growth in colon and gastric cancer cell lines. However, c9, t11-CLA was more potent than t10, c12-CLA in colon cell lines where both isomers were examined, though the optimal concentration level varied between studies (50 μmol/l and 200 μmol/l)⁽Reference Palombo, Ganguly and Bistrian^23, Reference Beppu, Hosokawa and Tanaka²⁴⁾. Subsequent work by Yasui et al. ⁽Reference Yasui, Suzuki and Kohno²⁵⁾ also confirmed the chemopreventive effect of c9, t11-CLA against pre-initiation (dose-dependent) as well as post-initiation stages of colorectal carcinogenesis (doses ≤  1% of diet).

Overall, in studies using animal models of cancer, the purified c9, t11-CLA isomer reduced tumorigenesis in six studies and showed no effect in two others⁽Reference Kelley, Hubbard and Erickson¹⁹⁾. Similarly, the t10, c12-CLA isomer decreased tumorigenesis in six studies, but in contrast increased tumorigenesis in two studies. Interestingly, three studies included in the present review found similar effects on the reduction of mammary tumours when a naturally enriched butter⁽Reference Ip, Banni and Angioni²⁶⁾ and synthetic isomers of c9, t11-CLA were fed to rats⁽Reference Lavillonniere, Chajes and Martin²⁷⁾ and mice⁽Reference Hubbard, Lim and Erickson²⁸⁾. Though more recent work suggests that t10, c12-CLA stimulates mammary tumours in a mouse model, where the gene erbB2/her2 is over-expressed, application of c9, t11-CLA showed no apparent effects⁽Reference Ip, McGee and Masso-Welch²⁹⁾. The same paper also demonstrated that the reduction in tumours was in the same order of magnitude irrespective of whether the CLA source was natural or synthetic. The authors of this paper suggest that it would be prudent to avoid supplements containing t10, c12-CLA in those at risk of developing breast cancer in which the erbB2/her2 gene is over-expressed (observed in 20–30% of human breast cancers), whereas supplements containing c9, t11-CLA may be safe and efficacious in breast cancer prevention⁽Reference Ip, McGee and Masso-Welch²⁹⁾. However, due to the differences in proliferation of tumours by the site of cancer, combining results may not elicit the true effects of CLA as an anti-carcinogenic agent, though in vitro and animal studies do demonstrate potential benefits.

The manifestation of cancer is not a practical end-point in human studies, combined with the numerous genetic and environmental risk factors for different cancers. Consequently, the majority of studies relating to CLA and cancer in humans are observational studies, particularly on breast cancer (Table 1). Dietary and serum CLA was shown to be significantly lower in postmenopausal cases of breast cancer compared with controls, thus suggesting a protective effect of CLA⁽Reference Aro, Mannisto and Salminen³⁰⁾. In a continuation of this study, breast adipose concentrations of CLA were not significantly different between cases and controls⁽Reference Chajes, Lavillonniere and Ferrari³¹⁾. Furthermore, there was no association between breast adipose tissue CLA concentration and prognostic factors of breast cancer or occurrence of metastases during a 7.5-year follow-up period⁽Reference Chajes, Lavillonniere and Maillard³²⁾. In the Netherlands Cohort Study on Diet and Cancer, intake of milk and milk products and meat products, as major sources of dietary CLA, showed no relationship with breast cancer incidence in postmenopausal patients⁽Reference Voorrips, Brants and Kardinaal³³⁾. This could be attributed to the fact that there were no significant differences in total CLA intake between cancer cases and controls⁽Reference Voorrips, Brants and Kardinaal³³⁾. The null association between breast cancer risk and intake of CLA was also demonstrated in a large epidemiological study in Sweden⁽Reference Larsson, Bergkvist and Wolk³⁴⁾. In contrast, in the same cohort, women who consumed four or more servings of high-fat dairy foods per d (including whole milk, full-fat cultured milk, cheese, cream, soured cream and butter) had a lower risk of developing colorectal cancer⁽Reference Larsson, Bergkvist and Wolk³⁵⁾. It has been suggested that a higher intake of c9, t11-CLA confers a reduced risk of a specific type of breast cancer tumour in premenopausal women. However, further investigation is warranted, as the sample size was small⁽Reference McCann, Ip and Ip³⁶⁾.

Table 1 Effect of conjugated linoleic acid (CLA) on cancer in human subjects

F, female; c9, t11, cis-9, trans-11; ER, oestrogen receptor; M, male; RCT, randomised controlled trial; t10, c12, trans-10, cis-12.

Recently, one small cross-over study examined colon cancer markers after subjects (n 15) consumed milk naturally enriched with c9, t11-CLA or synthetically enriched with t10, c12-CLA or normal milk as a control⁽Reference Farnworth, Chouinard and Jacques³⁷⁾. There were large variations in the responses to supplementation across all three groups (NS), therefore all data were combined and a significant decrease in enzyme activity β-glucosidase, nitroreductase and urease; P < 0·01 between day 0 and day 56 was observed. The authors stated that this was important due to links between enzymic activity and the production of carcinogens. However, it is important to note that the main aim of the study was to examine the effects of CLA-enriched milk on lipid metabolism and body composition⁽Reference Venkatramanan, Chouinard and Jacques³⁸⁾.

Currently the evidence for the anti-carcinogenic properties of CLA in human subjects is limited to observational studies, with broader epidemiological evidence not specifically focusing on CLA but rather on milk and dairy products. The World Cancer Research Fund & Association for International Cancer Research report reviewed the available evidence on the consumption of milk and links with cancer⁽³⁹⁾. The report concluded that milk probably protects against colorectal cancer, whereas there is limited evidence suggesting that cheese is a cause of colorectal cancer. There is also limited evidence suggesting that consumption of milk conveys a protective effect against bladder cancer. In contrast, diets high in Ca are a probable cause of prostate cancer, but there is limited evidence suggesting that high consumption of milk and dairy products is a cause of prostate cancer⁽³⁹⁾. Currently the evidence available is confusing, with suggestions that the effects are dependent on the site of the cancer due to the complex nature of diet, environment and nutrient interactions. However, a substantial amount of further work is required to fully elucidate the potential anti-carcinogenic properties of CLA in humans.

Conjugated linoleic acid and body composition

The overwhelming increases in the proportion of overweight and obesity in the world have been the focus of much debate and research. Currently two-thirds of the UK adult population are classified as overweight or obese (BMI > 25 kg/m²)⁽Reference Craig and Mindell⁴⁰⁾. Obesity is a multifaceted disorder, largely driven by its co-morbidities including type 2 diabetes, insulin resistance and CVD. To date feasible and sustainable approaches to prevent further increases in overweight and obesity, let alone attenuate it, have remained largely elusive. More recently, obesity has been recognised as a state of chronic or low-grade systemic inflammation, due to the abnormal circulating levels of inflammatory molecules, including TNFα, leptin and IL-6, which are secreted by adipose tissue⁽Reference Forsythe, Wallace and Livingstone⁴¹⁾.

Studies in animals have shown that feeding CLA at levels of 0.5–1% of the diet reduces body fat in mice, chickens, hamsters, rats and pigs⁽Reference Wahle, Heys and Rotondo⁴²⁾. The most substantial decreases in body fat have been observed in mice, where CLA at levels of 0.5% of the diet has been shown to lower body fat by 40 to 80%⁽Reference Wahle, Heys and Rotondo⁴²⁾. This effect is thought to be attributable to the t10, c12-CLA isomer, as the greatest body fat reductions in mice were observed with a CLA mix with a higher proportion of t10, c12-CLA than c9, t11-CLA⁽Reference Brown and McIntosh⁴³⁾. Also, in vitro, t10, c12-CLA prohibits TAG accumulation in cultures of differentiating human preadipocytes, whereas c9, t11-CLA increases TAG content⁽Reference Brown and McIntosh⁴³⁾. Evidence suggests that this effect may be due in part to a reduction in lipid uptake by adipocytes due to effects of CLA on stearoyl-CoA desaturase and lipoprotein lipase activity⁽Reference Pariza, Park and Cook⁴⁾.

Promising evidence from animal studies led to an array of human intervention studies being carried out investigating the effect of CLA on body composition in normal weight, overweight and obese subjects. The majority of these studies used a 50:50 (c9, t11-CLA and t10, c12-CLA) CLA mix, and results have been inconsistent. Almost all of these studies have shown no effect on body weight; however, some have reported reduced body fat mass (BFM) following supplementation with CLA, as discussed in detail below.

The body composition studies conducted in normal-weight adults (Table 2) have supplemented with 0.7–5.5 g 50:50 CLA/d, for 4–16 weeks, and of those which measured BFM, some have reported non-significant changes⁽Reference Petridou, Mougios and Sagredos^44, Reference Lambert, Goedecke and Bluett^46, Reference Nazare, de la Perriere and Bonnet^47, Reference Brown, Trenkle and Beitz^50, Reference Tricon, Burdge and Kew^79, Reference Kreider, Ferreira and Greenwood¹⁷⁵⁾, and others have reported BFM reductions of 4% up to 20%⁽Reference Mougios, Matsakas and Petridou^51–Reference Raff, Tholstrup and Toubro⁵⁶⁾. However, it is important to note that in some of the studies that have reported significant BFM reductions in normal-weight adults, subjects were involved in physical training throughout the supplementation periods, which may potentially be a confounder⁽Reference Thom, Wadstein and Gudmundsen^53–Reference Pinkoski, Chilibeck and Candow⁵⁵⁾.

Table 2 Effect of conjugated linoleic acid (CLA) on body composition in normal-weight human subjects

CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; M, male; F, female; RCT, randomised controlled trial; BFM, body fat mass; c9, t11, cis-9, trans-11; t10, c12, trans-10, cis-12; LBM, lean body mass.

* Subjects exercising.

In overweight and obese human subjects (Table 3), 50:50 CLA given at doses of 1.7–6.8 g/d, over periods of 4 to 104 weeks, has resulted in non-significant BFM changes in some instances⁽Reference Steck, Chalecki and Miller^57–Reference Joseph, Jacques and Plourde⁶³⁾, and reductions of 3–15% in other studies⁽Reference Brown, Trenkle and Beitz^50, Reference Steck, Chalecki and Miller^57–Reference Joseph, Jacques and Plourde^63, Reference Malpuech-Brugère, Verboeket-van de Venne and Mensink^77, Reference Risérus, Vessby and Arnlöv⁷⁸⁾. The greatest reduction in BFM (14.8%) was observed in a study of patients on blood pressure-lowering medication, who were supplemented with 4.5 g 50:50 CLA/d for 8 weeks⁽Reference Zhao, Zhai and Wang⁷¹⁾. In apparently healthy adults, the greatest reduction in BFM (6%) was observed in the study by Gaullier et al. ⁽Reference Gaullier, Halse and Hoye⁶⁶⁾ which was of 104 weeks' duration, and supplemented with 3.4 g 50:50 CLA/d. One study in children found that body fat gain was attenuated during prepubertal growth in 6–10-year-olds supplemented with 3.0 g 50:50 CLA/d⁽Reference Racine, Watras and Carrel⁷³⁾. However, in a few cases it has been noted that the largest reduction in BFM occurs in the lower body (for example, legs)⁽Reference Raff, Tholstrup and Toubro^56, Reference Gaullier, Halse and Hoivik⁶⁷⁾. Furthermore, some studies have reported increases in lean body mass (LBM) with CLA supplementation⁽Reference Steck, Chalecki and Miller^57, Reference Blankson, Stakkestad and Fagertun^64, Reference Gaullier, Halse and Hoivik⁶⁷⁾. In the study by Blankson et al. ⁽Reference Blankson, Stakkestad and Fagertun⁶⁴⁾ increased LBM was only observed in the group which significantly increased their level of intensive physical training during the intervention, hence it is possible that the observed effects were, at least partially, due to increased physical activity and not CLA supplementation.

Table 3 Effect of conjugated linoleic acid (CLA) on body weight or body composition in overweight and obese human subjects

CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; F, female; M, male; RCT, randomised controlled trial; BFM, body fat mass; SAD, sagittal abdominal diameter; t10, c12, trans-10, cis-12; c9, t11, cis-9, trans-11; LBM, lean body mass.

Interestingly, in another study, overweight subjects receiving 3.2 g of 50:50 CLA per d over a 6-month period, including the Christmas period, demonstrated a lower rate of weight gain and a 4% reduction in BFM compared with control⁽Reference Watras, Buchholz and Close⁶⁹⁾. A study of subjects with type 2 diabetes supplemented with 6 g of 50:50 CLA per d for 8 weeks found that plasma concentration of t10, c12-CLA, but not c9, t11-CLA, was inversely associated with body weight, suggesting that t10, c12-CLA is the active CLA isomer in relation to weight change⁽Reference Belury, Mahon and Banni⁷⁴⁾. This is in agreement with evidence from animal studies which also points to the t10, c12-CLA isomer as being the CLA isomer which elicits BFM reductions. A meta-analysis concluded that CLA, at a dose of 3.2 g/d, produces a modest body fat loss in humans of about 0.09 kg/week, with the relationship being linear up to 6 months⁽Reference Whigham, Watras and Schoeller⁷⁵⁾. This may be partly explained by the isomer- and tissue-specific effects of CLA, whereby c9, t11-CLA was found to be increased in skeletal muscle and t10, c12-CLA was incorporated into adipose tissue TAG in a subset of healthy non-obese participants⁽Reference Goedecke, Rae and Smuts⁷⁶⁾.

In addition to studies examining effects of CLA mixes, a number of studies have investigated the effects of individual CLA isomers on body composition. Findings from these studies show that consumption of 0.59–3 g c9, t11-CLA per d or 0.6–3.4 g t10, c12-CLA per d does not change body composition⁽Reference Risérus, Arner and Brismar^60, Reference Malpuech-Brugère, Verboeket-van de Venne and Mensink^77–Reference Tricon, Burdge and Kew⁷⁹⁾.

Currently, only three studies have been carried out which have fed subjects naturally CLA-enriched dairy products and investigated the effects on body composition⁽Reference Tricon, Burdge and Jones^45, Reference Brown, Trenkle and Beitz^50, Reference Desroches, Chouinard and Galibois⁸⁰⁾. In the study by Desroches et al. ⁽Reference Desroches, Chouinard and Galibois⁸⁰⁾, sixteen normolipidaemic overweight and obese men consumed butter naturally enriched with CLA (c9, t11-CLA; 2.59 g/d), or non-enriched control butter (0.24 g/d), for 4 weeks each in a cross-over design, and results showed no changes in body composition. Tricon et al. ⁽Reference Tricon, Burdge and Jones⁴⁵⁾ fed thirty-two healthy normolipidaemic men either naturally CLA-enriched or control dairy products (UHT full-fat milk, butter and cheese (1.42 v. 0.15 g c9, t11-CLA/d) in a 6-week cross-over study. Similarly, no changes in body weight were observed; however, body composition was not the primary outcome of this study, but rather blood lipid profile. No changes in body composition were observed when subjects consumed beef and dairy products naturally enriched with 1.17 g CLA/d for 56 d⁽Reference Brown, Trenkle and Beitz⁵⁰⁾. Also, with all of these studies it is important to note that the durations (4–8 weeks) were relatively short for investigating effects on body composition.

There are many possible explanations for the lack of reproducibility in studies of CLA's effect on body composition between animals and humans. These include age, sex, genetic predisposition to fat accumulation and differences in experimental design⁽Reference Plourde, Jew and Cunnane⁸¹⁾. It is interesting to note that although animal studies have evaluated the effects of CLA on weight gain over time in growing animals, the majority of human studies tend to investigate whether CLA affects weight or fat loss only in adults.

Conjugated linoleic acid, lipid metabolism and atherosclerosis

CVD are the leading cause of mortality globally⁽⁸²⁾ and so modification of key risk factors such as LDL-cholesterol or blood TAG are key targets (for example, in the UK⁽⁸³⁾). The impact of dietary fat and specific fatty acids on blood lipids has been a focus of research at least since Keys et al.'s early epidemiological work⁽Reference Keys, Aravanis and Blackburn⁸⁴⁾, so it is not surprising that the effect of CLA on blood lipids has been investigated.

Evidence from animal studies in rabbits, hamsters and mice has suggested that CLA has the potential to modulate plasma lipid metabolism and make an impact on the development and regression of cholesterol-induced atherosclerotic plaques⁽Reference Mitchell and McLeod⁸⁵⁾.

In rabbits, mixed-isomer CLA, fed at levels of 0.1–1% of diet over periods of 13 to 22 weeks, has been shown to reduce cholesterol deposition in the aorta⁽Reference Lee, Kritchevsky and Pariza⁸⁶⁾ and result in significant regression of established atherosclerotic lesions⁽Reference Kritchevsky, Tepper and Wright⁸⁷⁾. Furthermore, mixed-isomer CLA at a lower dose (0.05%) has been shown to be sufficient to decrease lesion development in rabbits⁽Reference Kritchevsky, Tepper and Wright⁸⁸⁾. Supplementation with either c9, t11-CLA or t10, c12-CLA results in similar reductions in lesion development to that seen with mixed-CLA isomer supplementation⁽Reference Kritchevsky, Tepper and Wright⁸⁹⁾.

Studies in hamsters that have supplemented with CLA over periods of 6–12 weeks, using different CLA isomers and doses, have shown mixed results, but there is evidence of improvements in lipid profile⁽Reference Nicolosi, Rogers and Kritchevsky^90, Reference Gavino, Gavino and Leblanc⁹¹⁾. In addition, there is some indication that CLA in conjunction with a lower-fat diet may reduce atherosclerotic lesions in the hamster⁽Reference Mitchell and McLeod⁸⁵⁾. It has been suggested that t10, c12-CLA may be the protective isomer in relation to lipid profile, as in the study by Gavino et al. ⁽Reference Gavino, Gavino and Leblanc⁹¹⁾, a CLA mix, but not the c9, t11-CLA isomer, improved the lipid profile of hamsters.

In mice, studies with supplemental CLA carried out over periods of 4–20 weeks, using different CLA isomers and doses, have also shown mixed results⁽Reference Mitchell and McLeod⁸⁵⁾. There has been one promising report of CLA (80:20 blend of c9, t11-CLA and t10, c12-CLA) resulting in marked regression of atherosclerotic lesions in apoE mice⁽Reference Toomey, Harhen and Roche⁹²⁾. In addition, there is some evidence of opposing effects of CLA isomers, with one study in mice showing c9, t11-CLA decreasing and t10, c12-CLA increasing atherosclerotic lesion area⁽Reference Arbones-Mainar, Navarro and Guzman⁹³⁾.

Further to the above studies which have supplemented animals' diets with commercial CLA preparations, studies have been carried out to investigate the anti-atherogenic effects of inclusion of dairy foods, and other foods such as eggs, naturally enriched with CLA, into the diets of animals⁽Reference Lock, Horne and Bauman^94–Reference Franczyk-Żarów, Kostogrys and Szymczyk⁹⁸⁾. The results of these studies have shown that CLA can improve plasma lipid profile and decrease atherosclerosis-related biomarkers. Overall, at present there is no general consensus as to the effect of CLA supplementation on lipids or atherosclerosis in animals. Furthermore, most animal studies that have suggested protective anti-atherogenic effects have generally provided CLA doses greater than those achievable in the human diet.

Despite much investigation, the precise mechanisms by which CLA affects lipid metabolism and adipose tissue are not fully elucidated. However, it is thought that CLA modulates energy expenditure, apoptosis, fatty acid oxidation, lipolysis and lipogenesis⁽Reference House, Cassady and Eisen⁹⁹⁾. As discussed in the previous section, the t10, c12-CLA isomer is thought to exert effects on body composition, partly due to a reduction in lipid uptake by adipocytes due to effects of CLA on stearoyl-CoA desaturase and lipoprotein lipase activity⁽Reference Pariza, Park and Cook⁴⁾.

In humans epidemiological studies on dietary CLA intakes and prevalence of atherosclerosis have not been carried out to date. However, over the past decade, numerous human intervention studies have investigated the effect of CLA on lipids and other markers of atherosclerotic risk (Table 4), the results of which have been highly inconsistent, possibly due to the use of different isomers and varying doses. The majority of these studies have used commercial mixed- or pure-isomer CLA preparations, at levels of 1.7 to 6.8 g/d, over periods of 4 to 13 weeks, and have not shown any overall effect on plasma lipid or lipoprotein concentrations, compared with placebo, in normal-weight and overweight subjects⁽Reference Petridou, Mougios and Sagredos^44, Reference Lambert, Goedecke and Bluett^46, Reference Brown, Trenkle and Beitz^50, Reference Smedman and Vessby^52, Reference Herrmann, Rubin and Häsler^61–Reference Blankson, Stakkestad and Fagertun^64, Reference Racine, Watras and Carrel^73, Reference Risérus, Vessby and Arnlöv^78, Reference Benito, Nelson and Kelley^100–Reference Engberink, Geleijnse and Wanders¹⁰⁵⁾. However, one study did report significant within-group reductions in total cholesterol and LDL-cholesterol with doses of 1.7 and 3.4 g CLA/d⁽Reference Blankson, Stakkestad and Fagertun⁶⁴⁾, but it was stated that the reductions were not clinically important.

Table 4 Effect of conjugated linoleic acid (CLA) on blood lipid concentrations in human subjects

CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; M, male; F, female; RCT, randomised controlled trial; tChol, total cholesterol; c9, t11, cis-9, trans-11; t10, c12, trans-10, cis-12.

* Same study with results reported over two papers.

Some studies have reported that supplementation with commercial CLA preparations can have a negative effect on the lipid profile. For example, a significant decrease in HDL-cholesterol was observed on supplementing with 3.4 g t10, c12-CLA per d in obese men with the metabolic syndrome⁽Reference Risérus, Arner and Brismar⁶⁰⁾, and in healthy subjects who were supplementing their diets with 0.7–1.4 g CLA mix per d⁽Reference Mougios, Matsakas and Petridou⁵¹⁾. There is some evidence to suggest that CLA (mixtures and individual isomers) can induce lipid peroxidation⁽Reference Risérus, Vessby and Arnlöv^78, Reference Raff, Tholstrup and Basu¹⁰³⁾; however, it is not known whether this effect of CLA could be pro-atherogenic in humans.

In contrast, other studies have shown a positive effect of CLA, with 3 g 50:50 CLA mix per d lowering fasting TAG, and 3 g 80:20 CLA mix per d decreasing VLDL, in healthy subjects⁽Reference Noone, Roche and Nugent¹⁰⁶⁾. Furthermore, 3 g 50:50 CLA per d was shown to significantly increase HDL-cholesterol and significantly decrease LDL:HDL-cholesterol in patients with type 2 diabetes⁽Reference Moloney, Yeow and Mullen¹⁰⁷⁾. Consumption of foods enriched with 26.8 g CLA/d led to a significant positive effect on HDL concentration and a significant lowering of LDL-cholesterol⁽Reference Wanders, Brouwer and Siebelink⁴⁹⁾. Interestingly, Tricon et al. ⁽Reference Tricon, Burdge and Kew⁷⁹⁾ observed divergent responses in plasma lipids with CLA supplementation, with t10, c12-CLA (0.6–2.5 g/d) increasing LDL:HDL-cholesterol and total:HDL-cholesterol and c9, t11-CLA (0.59–2.38 g/d) decreasing these ratios, with no dose-dependent effect observed. Elevated cholesterol ratios of LDL:HDL and total:HDL-cholesterol are independent risk factors for CHD⁽Reference Ridker, Stampfer and Rifai^108, Reference Lewington, Whitlock and Clarke¹⁰⁹⁾.

Recently, the effects of consuming dairy products, naturally rich in CLA or naturally enriched with CLA, on lipids in human subjects have been examined in four studies⁽Reference Tricon, Burdge and Jones^45, Reference Brown, Trenkle and Beitz^50, Reference Desroches, Chouinard and Galibois^80, Reference Sofi, Buccioni and Cesari¹¹⁰⁾. Three of these studies manipulated cows' diets to produce dairy products naturally enriched with CLA⁽Reference Petridou, Mougios and Sagredos^44, Reference Brown, Trenkle and Beitz^50, Reference Desroches, Chouinard and Galibois⁸⁰⁾. In the study by Desroches et al. ⁽Reference Desroches, Chouinard and Galibois⁸⁰⁾, normolipidaemic overweight and obese men consumed butter naturally enriched with CLA (c9, t11-CLA; 2.59 g/d), or non-enriched control butter (0.24 g/d), for 4 weeks. Results showed plasma lipid subfraction levels (VLDL, LDL and HDL) were not significantly different between the two treatments; however, consumption of the non-enriched butter resulted in a significantly greater reduction of total cholesterol, total:HDL-cholesterol and LDL:HDL-cholesterol compared with the CLA-enriched butter, a result which was contradictory to the hypothesis. Tricon et al. ⁽Reference Tricon, Burdge and Jones⁴⁵⁾ fed healthy normolipidaemic men either naturally CLA-enriched or control dairy products (UHT full-fat milk, butter and cheese (1.42 v. 0.15 g c9, t11-CLA per d)) in a 6-week cross-over study. Overall, lipid subfractions were not affected; however, a small but significant increase in LDL:HDL-cholesterol was observed. These results were similar to findings by Brown et al. ⁽Reference Brown, Trenkle and Beitz⁵⁰⁾ where consumption of beef and dairy products rich in CLA (1.17 g CLA/d) for 56 d did not alter blood lipid profile.

A small, cross-over study in ten healthy subjects found that consumption of cheese made from naturally CLA-rich sheep's milk (0.25 g c9, t11-CLA per d) for 10 weeks had no effect on plasma lipids, as compared with consumption of a regular cows' cheese⁽Reference Sofi, Buccioni and Cesari¹¹⁰⁾. The daily intake of CLA was confirmed as being 0.25 g c9, t11-CLA in correspondence with the author. It is important to note that using cows' milk cheese as a control was not ideal, due to the fact that it has a very different fatty acid profile compared with sheep's cheese. Overall these studies have shown no significant effect of treatment with dairy products naturally rich in CLA or naturally enriched with CLA on plasma lipids.

Dairy products which are naturally enriched in CLA are also higher in trans-vaccenic acid (trans-18 : 1), lower in SFA content, and slightly higher in n-3 PUFA content than conventional dairy products, due to the feeding strategies employed for enrichment⁽Reference Jones, Shingfield and Kohen¹⁵⁾. It has been suggested that consuming trans-fatty acids impairs the lipid profile by lowering HDL-cholesterol and raising LDL-cholesterol levels⁽Reference Lichtenstein, Ausman and Jalbert¹¹¹⁾. Whether the content of trans-vaccenic acid in naturally CLA-enriched dairy products could counteract the potential benefit of CLA on the lipid profile unclear. The current evidence examining the intake of trans-fatty acids from animal sources and associations with CHD presents a confusing picture, particularly given the higher than typically consumed levels of trans-fatty acids used within studies⁽Reference Willett, Stampfer and Manson^112–Reference Mensink¹¹⁶⁾. However, it is unclear whether the partial conversion of trans-vaccenic to c9, t11-CLA in human intestines, liver and adipose tissue promotes adverse or beneficial effects on lipid profile⁽Reference Mozaffarian, Katan and Ascherio^113, Reference Turpeinen, Mutanen and Aro¹¹⁷⁾.

The reason for the inconsistent and mostly neutral results in relation to the effects of CLA on lipids in human studies compared with animal studies is unclear. However, it is important to note that while animal studies examined the effect of using CLA to supplement hyperlipidaemic animals that were eating atherogenic diets, human studies examined the effect of supplementing diets of normolipidaemic subjects with CLA. Furthermore, it is conceivable that the anti-atherosclerotic effects of CLA observed in animal studies may be due to mechanisms other than effects on lipids, for instance anti-inflammatory effects, as atherosclerosis is an inflammatory disease.

Conjugated linoleic acid, inflammation and immune effects

Inflammation underlies a wide range of conditions. For example, as noted above, obesity is now recognised as a state of chronic or low-grade systemic inflammation, due to the abnormal circulating levels of inflammatory molecules, including TNFα, leptin and IL-6, which are secreted by adipose tissue⁽Reference Forsythe, Wallace and Livingstone⁴¹⁾. In addition, inflammation is central to atherosclerosis⁽Reference Libby, Ridker and Hansson¹¹⁸⁾ and the metabolic syndrome⁽Reference Gade, Schmit and Collins¹¹⁹⁾.

In vitro studies have shown that CLA has anti-inflammatory effects. CLA (CLA mix, or c9, t11-CLA or t10, c12-CLA) is associated with a lower mRNA expression of the inflammatory mediators cyclo-oxygenase-2, TNFα, and inducible NO synthase, and decreases production of induced PGE₂, NO, IL-6 and IL-1β in mouse macrophage cells⁽Reference Yu, Correll and Vanden Heuvel¹²⁰⁾. The c9, t11-CLA isomer inhibits induced eosinophil activation, decreases transcription of TNFα, IL-6 and IL-12 in Caco-2 cells and enhances IL-10 production in murine dendritic cells⁽Reference Jaudszus, Foerster and Kroegel^121–Reference Reynolds, Loscher and Moloney¹²³⁾. Furthermore, both c9, t11-CLA and t10, c12-CLA reduce PGE₂ and thromboxane B₂ concentrations in human macrophages⁽Reference Stachowska, Dolegowska and Dziedziejko¹²⁴⁾.

Animal studies have been carried out to determine if CLA exerts anti-inflammatory effects in vivo; however, results to date have been inconsistent. Three animal studies have found a CLA mix to be anti-inflammatory⁽Reference Changhua, Jindong and Defa^125–Reference Noto, Zahradka and Ryz¹²⁷⁾. Obese rats fed 1.5% CLA mix for 8 weeks were found to have less adipose TNFα mRNA expression; however, other markers of inflammation did not change⁽Reference Noto, Zahradka and Ryz¹²⁷⁾. Butz et al. ⁽Reference Butz, Li and Huebner¹²⁶⁾ reported that mice fed 0.5% CLA mix for 3 weeks had less plasma TNFα compared with mice on a control diet. In pigs fed 2% CLA mix for 14 d, a decrease in induced elevation and mRNA expression of pro-inflammatory cytokines (IL-6 and TNF-α), and an increase in an anti-inflammatory cytokine (IL-10) were observed. Furthermore, a molecular aspect of the same study determined t10, c12-CLA to be the main isomer to which the anti-inflammatory effect can be attributed⁽Reference Changhua, Jindong and Defa¹²⁵⁾. However, in contrast to these findings, two studies have established t10, c12-CLA to have pro-inflammatory effects, where mice fed 0.5% t10, c12-CLA for 14 d showed induced pro-inflammatory cytokine transcripts in white adipose tissue⁽Reference LaRosa, Miner and Xia¹²⁸⁾, and short-term supplementation with t10, c12-CLA in mice also increased pro-inflammatory cytokine gene expression in a study⁽Reference Poirier, Shapiro and Kim¹²⁹⁾.

Human intervention studies have investigated the effect of CLA (both commercial preparations and naturally CLA-enriched dairy products) on various biomarkers of inflammation (Table 5). Results to date have been mixed, with most studies either showing an increase in inflammatory markers, or no change. Three studies that have supplemented subjects with a CLA mixture at doses of 4.2 to 6.4 g/d, over periods of 12 to 16 weeks, have found increases in plasma levels of C-reactive protein (CRP)⁽Reference Steck, Chalecki and Miller^57, Reference Smedman, Basu and Jovinge^130, Reference Tholstrup, Raff and Straarup¹³¹⁾. There were no significant effects on inflammatory markers including CRP and a range of interleukins when subjects were supplemented with 4 to 4.5 g CLA mixture/d⁽Reference MacRedmond, Singhera and Attridge^132, Reference Stickford, Mickleborough and Fly¹³³⁾. Two studies with CLA added to foods showed no effect on plasma CRP levels; however, the duration of these trials was relatively short (5 and 8 weeks)⁽Reference Joseph, Jacques and Plourde^63, Reference Raff, Tholstrup and Basu¹⁰³⁾. Furthermore, two crossover studies that provided c9, t11-CLA at doses of 4 g/d⁽Reference Sluijs, Plantinga and de Roos⁶²⁾ or 0.6–2.4 g/d and 0.6–2.5 g/d t10, c12-CLA⁽Reference Tricon, Burdge and Kew¹³⁴⁾, for 6 months and 8 weeks respectively, observed no change in plasma CRP concentrations.

Table 5 Effect of conjugated linoleic acid (CLA) on inflammation and other immune indices in human subjects

CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; F, female; DTH, delayed-type hypersensitivity; t10, c12, trans-10, cis-12; M, male; RCT, randomised controlled trial; c9, t11, cis-9, trans-11; CRP, C-reactive protein; VCAM, circulating vascular adhesion molecule; PBMC, peripheral blood mononuclear cell; ICAM, intercellular adhesion molecule; LT, leucotriene; E-selectin, endothelial leucocyte adhesion molecule; PAI, plasminogen activator inhibitor; MCP, monocyte chemoattractant protein; FVII-C, factor VII coagulant; EDN, eosinophil-derived neurotoxin; GM-CSF, granulocyte macrophage colony-stimulating factor; IFN, interferon; ACE, angiotensin-converting enzyme; ECP, eosinophil cationic protein; LTC₄-E₄, cysteinyl 4-series leukotrienes.

Supplementation with t10, c12-CLA at doses of 3–3.4 g/d for 12–13 weeks has produced inconsistent results. A study in obese men with the metabolic syndrome found increased plasma CRP levels; on the other hand, a study in overweight men and women demonstrated no effect on plasma CRP, or on other markers of inflammation⁽Reference Ramakers, Plat and Sebedio^135, Reference Risérus, Basu and Jovinge¹³⁶⁾. In the case of c9, t11-CLA, supplementation with similar doses (3 g) for similar durations (12–13 weeks) has also resulted in contrasting results, with one study reporting increased excretion of a pro-inflammatory marker (15-keto-dihydro-PGF2α) in obese subjects⁽Reference Risérus, Vessby and Arnlöv⁷⁸⁾, and another study reporting no effect on a range of pro-inflammatory markers in overweight subjects⁽Reference Ramakers, Plat and Sebedio¹³⁵⁾.

As described in the previous section, the effect of feeding subjects dairy products which are naturally enriched in c9, t11-CLA (due to the manipulation of diets of cows) has been investigated in two studies to date⁽Reference Tricon, Burdge and Jones^45, Reference Desroches, Chouinard and Galibois⁸⁰⁾. In these studies, daily doses of 1.4–2.6 g c9, t11-CLA were fed for durations of 4–6 weeks, and no changes in plasma CRP concentrations and other inflammatory markers were observed. In contrast, a study by Sofi et al. ⁽Reference Sofi, Buccioni and Cesari¹¹⁰⁾ found that consumption of sheep cheese, naturally rich in CLA (0.25 g c9, t11-CLA per d), for 10 weeks decreased circulating levels of the pro-inflammatory cytokines IL-6, IL-8 and TNFα, compared with consumption of a control cows' cheese. However, as noted above, this study was small, poorly controlled and may not have been adequately powered for the multiple variables measured.

Some studies have investigated other immune effects in addition to inflammation. A study where the diets of young women were supplemented with a CLA mixture at 3.9 g/d for 9 weeks found that no indices of immune status were affected (such as the number of circulating leucocytes; granulocytes; monocytes; lymphocytes and their subsets; lymphocytes proliferation in response to phytohaemagglutinin and influenza vaccine; and serum influenza antibody titres)⁽Reference Kelley, Taylor and Rudolph¹³⁷⁾. However, the sample size was small, at seventeen. In a larger study, with fifty-five subjects, Nugent et al. ⁽Reference Nugent, Roche and Noone¹³⁸⁾ found that either a 50:50 CLA mixture or an 80:20 CLA mixture at about 2 g/d had minimal effects on lymphocytes and cytokines, and had no additional benefit on immune function compared with linoleic acid. CLA supplementation has also been linked to reduced symptoms of birch pollen allergy⁽Reference Turpeinen, Ylönen and von Willebrand¹⁰⁴⁾ and improved airway hyper-responsiveness in asthmatics⁽Reference MacRedmond, Singhera and Attridge¹³²⁾. However, a second study in asthmatics found no attenuation of airway inflammation or bronchoconstrictive response⁽Reference Stickford, Mickleborough and Fly¹³³⁾.

However, Song et al. ⁽Reference Song, Grant and Rotondo¹³⁹⁾ found that supplementing twenty-eight males and females with 3 g CLA 50:50 for 12 weeks had beneficial effects on immune function as it decreased pro-inflammatory cytokines (TNFα and IL-1β) and increased an anti-inflammatory cytokine (IL-10). Furthermore, Tricon et al. ⁽Reference Tricon, Burdge and Kew¹³⁴⁾ found that supplementing men with 0.6 to about 2.5 g of either c9, t11-CLA or t10, c12-CLA per d decreased mitogen-induced T lymphocyte activation dose-dependently (however, lymphocytes and cytokines were unaffected). Mullen et al. ⁽Reference Mullen, Moloney and Nugent¹⁴⁰⁾ showed that 2.2 g CLA 50:50 per d for 8 weeks decreased stimulated peripheral blood mononuclear cell IL-2 secretion, but did not affect other markers including plasma levels of IL-6, CRP, fibrinogen or TNFα, in thirty men.

Overall, studies investigating the effect of CLA (both supplements and naturally CLA-enriched products) on immune indices and inflammation provide inconsistent results.

Conjugated linoleic acid, insulin resistance and diabetes

In addition to the potential anti-atherogenic, anti-obesity and anti-inflammatory properties of CLA, the effects on diabetes have also been examined. As previously stated, increases in overweight and obesity have been concurrent with increases in type 2 diabetes, which is characterised by insulin resistance and occurs as a result of excess adipose tissue. A 5% reduction in body weight has been shown to decrease insulin resistance in overweight and obese subjects⁽Reference Brown and McIntosh^43, Reference Taylor and Zahradka^141, Reference Belury¹⁴²⁾. Therefore the observed modest reductions in body weight with CLA mixtures at 3 g/d may also improve insulin resistance.

Overall, the results from both animal and in vitro work are conflicting, with the effects of CLA on insulin resistance examined in addition to other outcomes (atherogenic and obesogenic properties). The vast majority of studies have examined the effects of CLA isomer mixtures, though some results do suggest isomeric differences⁽Reference Roche, Noone and Sewter¹⁴³⁾. In a mouse model, feeding a diet rich in t10, c12-CLA induced insulin resistance whereas c9, t11-CLA improved lipid metabolism without impairing insulin action⁽Reference Moloney, Toomey and Noone¹⁴⁴⁾ by possible mediation of the pro-inflammatory state⁽Reference Houseknecht, Vanden Heuvel and Moya-Camarena¹⁴⁵⁾. Similarly, studies with male Zucker diabetic fatty (ZDF) rats feeding a 50:50 blend of CLA reduced glucose and insulin concentrations⁽Reference Ryder, Portocarrero and Song¹⁴⁶⁾, although the diet with 91% c9, t11-CLA showed no effect⁽Reference Tsuboyama-Kasaoka, Takahashi and Tanemura¹⁴⁷⁾. In contrast, in another mouse model of diabetes, a blend of CLA isomers induced marked lipodystrophic insulin resistance and glucose tolerance⁽Reference Halade, Rahman and Fernandes^148, Reference Halade, Rahman and Fernandes¹⁴⁹⁾. In the same strain of young and ageing mice, supplementation with the individual isomers or a CLA mix demonstrated divergent responses⁽Reference Halade, Rahman and Fernandes^148, Reference Halade, Rahman and Fernandes¹⁴⁹⁾. Supplementation, with c9, t11-CLA elicited no effects on indices of insulin resistance, plasma insulin and glucose, whereas supplementation with t10, c12-CLA or a CLA mix increased plasma glucose, insulin and homeostasis model assessment of insulin resistance (HOMA-IR). However, during an intravenous glucose tolerance test, mice supplemented with c9, t11-CLA eliminated glucose faster than the control, t10, c12-CLA- or CLA mix-fed mice⁽Reference de Roos, Rucklidge and Reid¹⁵⁰⁾. These data highlight the importance of not just measuring plasma glucose and insulin, as true effects may only be apparent when more robust measures of insulin resistance are used.

One group has used a proteomics approach for eliciting the interactions between CLA isomers and diseases in an animal model⁽Reference de Roos and McArdle¹⁵¹⁾. Proteomic techniques measure changes in the protein complement of a biological system and enable modelling of biological processes in response to dietary interventions⁽Reference de Roos, Rucklidge and Reid¹⁵⁰⁾. In a study with apoE mice consuming 7% c9, t11-CLA or t10, c12-CLA or control (linoleic acid), results suggested that c9, t11-CLA exerted anti-diabetic effects due to altered expression of markers, whereas t10, c12-CLA asserted pro-diabetic effects⁽Reference Risérus, Arner and Brismar^60, Reference Risérus, Vessby and Arnlöv^78, Reference Risérus, Smedman and Basu¹⁵²⁾. Overall, this study suggests that c9, t11-CLA potentially contributes to a less severe inflammatory response or protection against the development of atherosclerosis. However, conducting a trial in human subjects would be prohibitively expensive and require a rigorously controlled protocol in order to examine the effects of CLA supplementation on protein structure and function.

Currently the anti-diabetic properties of CLA in human subjects (Table 6) cannot be fully determined, as few studies are undertaken using rigorous measures of insulin resistance such as the hyperinsulinaemic–euglycaemic clamp⁽Reference Moloney, Yeow and Mullen^107, Reference Ahren, Mari and Fyfe¹⁵³⁾ or the oral glucose tolerance test⁽Reference Brown, Trenkle and Beitz^50, Reference Raff, Tholstrup and Toubro^56, Reference Syvertsen, Halse and Hoivik^58, Reference Gaullier, Halse and Hoye^65–Reference Gaullier, Halse and Hoivik^67, Reference Noone, Roche and Nugent^106, Reference Tarnopolsky, Zimmer and Paikin^154, Reference Eyjolfson, Spriet and Dyck¹⁵⁵⁾. Indeed the majority of results on the anti-diabetic properties of CLA relate to studies where only fasting plasma or serum glucose or insulin have been measured, are not the main focus of the study and typically have small sample sizes. Given these limitations, it is perhaps not surprising that the overall results show no effects of CLA supplementation⁽Reference Raff, Tholstrup and Toubro^56, Reference Herrmann, Rubin and Häsler^61, Reference Sluijs, Plantinga and de Roos^62, Reference Gaullier, Halse and Hoye^65–Reference Gaullier, Halse and Hoivik^67, Reference Norris, Collene and Asp^70, Reference Zhao, Zhai and Wang^71, Reference Turpeinen, Ylönen and von Willebrand^104, Reference Noone, Roche and Nugent^106, Reference MacRedmond, Singhera and Attridge^132, Reference Tarnopolsky, Zimmer and Paikin¹⁵⁴⁾ or consumption of CLA-enriched products⁽Reference Tricon, Burdge and Jones^45, Reference Joseph, Jacques and Plourde^63, Reference Racine, Watras and Carrel^73, Reference Naumann, Carpentier and Saebo^102, Reference Raff, Tholstrup and Basu¹⁰³⁾ on glucose and insulin. However, supplementing with a CLA mixture has shown beneficial effects on insulin resistance in healthy male subjects⁽Reference Eyjolfson, Spriet and Dyck¹⁵⁵⁾ and type 2 diabetic subjects⁽Reference Belury, Mahon and Banni⁷⁴⁾. In contrast, a negative effect on insulin resistance was reported in type 2 diabetic patients; however, this may have been due to the bias in the glucose tolerance between the supplementation and placebo groups and may not have been due to CLA supplementation⁽Reference Moloney, Toomey and Noone¹⁴⁴⁾.

Table 6 Effect of conjugated linoleic acid (CLA) on insulin resistance in human subjects

c9, t11, cis-9, trans-11; t10, c12, trans-10, cis-12; M, male; F, female; RCT, randomised controlled trial; OGTT, oral glucose tolerance test; CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; EGC, hyperinsulinaemic-euglycaemic clamp; HOMA-IR, homeostasis model assessment of insulin resistance; QUICKI, quantitative insulin sensitivity check index.

A recent study also found increased insulin resistance in older obese subjects, but no effects of combined CLA–n-3 supplementation in lean or obese younger subjects or older lean subjects⁽Reference Ahren, Mari and Fyfe¹⁵³⁾. Supplementation with the individual isomers, c9, t11-CLA or t10, c12-CLA increased insulin resistance (+15%) in obese men with the metabolic syndrome⁽Reference Risérus, Arner and Brismar^60, Reference Risérus, Vessby and Arnlöv⁷⁸⁾, whereas a CLA isomer mixture did not affect insulin resistance⁽Reference Risérus, Arner and Brismar⁶⁰⁾. Furthermore, lipid peroxidation increased relative to placebo when the individual isomers were administered, but the differences did not remain significant when adjusted for changes in lipid peroxidation⁽Reference Risérus, Arner and Brismar^60, Reference Risérus, Vessby and Arnlöv⁷⁸⁾. The authors of these papers suggest that irrespective of the CLA isomer, CLA-induced lipid peroxidation may mediate insulin resistance. However, further work is required, particularly studies where the hyperinsulinaemic–euglycaemic clamp is utilised⁽Reference Risérus, Smedman and Basu^152, Reference Risérus¹⁵⁶⁾. The conflicting responses to increased CLA intake in both human and animal studies do not currently imply compelling anti-diabetic properties of CLA. Thus, studies should be designed that provide rigorous measures of insulin resistance in subjects of varying age groups and weight status⁽Reference Alberti, Eckel and Grundy¹⁵⁷⁾.

Conjugated linoleic acid and bone health

Bone is a complex tissue system whereby the skeleton is continually renewed through the resorption (breakdown) of existing bone and the formation of new bone (remodelling). Peak bone mass in humans usually occurs late in the second or early in the third decade of life with a progressive decline in bone mineral density starting in the fourth decade of life for both men and women⁽Reference Prentice, Schoenmakers and Laskey¹⁵⁸⁾. Bone modelling (children and young adults) or remodelling (adults) is influenced by many factors including nutritional status, hormones and mechanical loading. One of the consequences of low bone turnover or remodelling is the development of osteoporosis, particularly in white, postmenopausal women. In the UK, the costs of osteoporosis to the National Health Service are estimated at £2.3 billion per year or £6 million per d, with almost 3 million individuals diagnosed with osteoporosis⁽¹⁵⁹⁾. Thus, strategies that attenuate decreases in bone mass are of great importance, with much of research focused on Ca, vitamin D, protein and vitamin K intakes⁽Reference Lanham-New¹⁶⁰⁾. However, other nutrients, including CLA, have been the focus of research due to influences on bone mass and metabolism⁽Reference Bhattacharya, Banu and Rahman^18, Reference Roy and Antolic^161–Reference Watkins and Seifert¹⁶³⁾.

The majority of work on CLA and bone metabolism has been conducted using human cells and animal models, particularly those reflecting postmenopausal women. Supplementation studies have demonstrated decreased PGE₂ production in rats, but results were dependent on the CLA concentration levels⁽Reference Kelly and Cashman^164–Reference Watkins, Shen and McMurtry¹⁶⁸⁾. PGE₂ is an important factor in the regulation of bone metabolism, including bone formation as well as bone resorption⁽Reference Prentice, Schoenmakers and Laskey¹⁵⁸⁾. PGE₂ production increases in postmenopausal bone loss due to oestrogen deficiency⁽Reference Prentice, Schoenmakers and Laskey¹⁵⁸⁾. CLA may also stimulate Ca absorption, thus making more Ca available for bone formation⁽Reference Kelly and Cashman^164, Reference Jewell and Cashman¹⁶⁹⁾. Recently, Park et al. ⁽Reference Park, Pariza and Park¹⁷⁰⁾ reanalysed previous studies in mice and showed that extra Ca (0.66%) in the diet improved CLA effects on bone mass in male, but not female mice. A recent review concluded that based on the current evidence from in vitro and animal studies the addition of CLA, overall, improves bone strength and density⁽Reference Roy and Antolic¹⁶¹⁾. However, the majority of studies currently published were conducted using CLA isomer mixtures. Only two studies have examined the differences between the c9, t11 and t10, c12 isomers and found no direct effects on bone, but rather attenuation of parathyroid hormone concentration⁽Reference Weiler, Austin and Fitzpatrick-Wong^171, Reference Weiler, Fitzpatrick and Fitzpatrick-Wong¹⁷²⁾.

Whilst there are numerous publications examining the effects of CLA and bone formation in cell and animal models, studies in human subjects are lacking (Table 7). Data from an observational study showed that in postmenopausal women dietary intake of CLA was a weak but significant predictor of Ward's triangle bone mineral density⁽Reference Brownbill, Petrosian and Ilich¹⁷³⁾. The same study also found that subjects with above median intake of CLA had higher bone mineral density of the forearm. In contrast, supplementation with a CLA mix (3.0–3.4 g/d) did not affect bone formation or resorption in healthy lean, overweight, obese men and women⁽Reference Gaullier, Halse and Hoye^65–Reference Gaullier, Halse and Hoivik^67, Reference Doyle, Jewell and Mullen¹⁷⁴⁾. A further two studies in young and elderly subjects who completed resistance training in addition to CLA supplementation (6 g/d) also demonstrated no change in bone mineral density and bone mass⁽Reference Tarnopolsky, Zimmer and Paikin^154, Reference Kreider, Ferreira and Greenwood¹⁷⁵⁾. Brown et al. ⁽Reference Brown, Trenkle and Beitz⁵⁰⁾ reported no change in bone mineral content when subjects consumed a CLA-enriched diet, although the study duration was only 56 d, an insufficient length of time for observing changes in bone mineral content. In children, significantly less bone mineral content accretion occurred in the CLA supplemented after 7 months⁽Reference Racine, Watras and Carrel⁷³⁾; however, the reasons are not fully elucidated. Currently, the only human study to demonstrate a positive association between CLA supplementation(5 g/d) and bone found a decrease in bone resorption markers and increase in LBM⁽Reference Pinkoski, Chilibeck and Candow⁵⁵⁾. However, this study did not identify whether the increases in LBM were due to increased muscle or bone mass and whether it was an artifact of the 7-week resistance training programme. Since there are relatively few human studies (four out of seven studies where bone was not the primary outcome examined), the lack of consistency in protocols, measurement of bone metabolites and small sample sizes hinder a clear conclusion between the effects of CLA and bone.

Table 7 Effect of conjugated linoleic acid (CLA) on bone health in human subjects

CLA mixture, 50:50 cis-9, trans-11- and trans-10, cis-12-CLA; M, male; RCT, randomised controlled trial; BMD, bone mineral density; F, female; BMC, bone mineral content; OGTT, oral glucose tolerance test.

Overall conclusions

The overall evidence from the studies examined here demonstrates a lack of definitive and reproducible results, particularly in relation to the consumption of naturally enriched CLA products, as the number of published studies is low relative to the number on synthetic supplements. The majority of randomised controlled trials are conducted with CLA supplements, with varying mixtures of isomers and dosage levels. However, the evidence from animal studies is promising, but extrapolation from animal to human studies is difficult due to the differences in the amount of CLA used. For example, in animal studies the observed benefits of CLA on bone are between 0.1–1% CLA of total weight of diet⁽Reference Brownbill, Petrosian and Ilich¹⁷³⁾. For men consuming on average 3.0 kg food and beverages per d, this is equivalent to 3–30 g CLA/d; for women consuming about 2.2 kg food and beverages per d, this equates to 2.2–22 g CLA/d⁽Reference Kant and Graubard¹⁷⁶⁾. In addition, given the differences in study protocols, relatively small sample sizes and other methodological issues (including measurement of dietary CLA intakes⁽Reference Ritzenthaler, McGuire and Falen⁸⁾ and accurate measurement of body composition), it is not surprising that there is a lack of consensus on what health claims could be applicable to CLA, either natural or synthetic products. Current submissions on CLA health claims to the European Food Safety Authority (EFSA) include seven for body weight/LBM, two on immune function, two on antioxidant properties and one relating to insulin. The present review suggests that the only possible candidate would be in relation to the synthetic t10, c12-CLA isomer and reductions in body fat.