The incidence of obesity and overweight in the US has increased considerably during the past two decades and currently affects 65% of the adult population. Research has indicated that small, yet irreversible, gains during the holiday season contribute to increases in weight during adulthood. Conjugated linoleic acid (CLA), a naturally occurring dietary fatty acid, has been found to reduce weight gain and dramatically decrease fat mass in animals. Although research in humans has shown inconsistent results, most studies have been of insufficient duration or have utilized body composition methods that are less accurate than the currently accepted criterion.
Randomized, double-blind, placebo-controlled study of 3.2 g/day CLA for 6 months.
Forty healthy, overweight subjects (age: 18–44 years; body mass index: 25–30 kg/m2)
Body composition by the four-compartment model, resting metabolic rate (RMR) by indirect calorimetry, self-reported physical activity and dietary intake, and blood chemistries were determined at baseline and after 6 months. Body weight was measured monthly during the pre-holiday season (August–October), holiday season (November–December) and post-holiday season (January–March). Adverse events were assessed monthly.
Compared to CLA, the placebo group showed a greater rate of weight gain during the holiday season (P=0.01). Within the placebo group, holiday weight change was significantly greater compared to the pre-holiday period (August–October) (P=0.03). Six-month change in body composition was improved with CLA compared to placebo (P=0.02), and body fat was significantly reduced within the CLA group (−1.0±2.2 kg, P=0.05). CLA had no effect on RMR, physical activity or dietary intake. The rate of reported negative emotions decreased significantly with CLA, although there was no difference in any other category of adverse event. In comparison to the placebo, CLA did not affect insulin resistance, blood lipids and markers of liver function or markers of inflammation, with the exception of a significant decrease in a biomarker of endothelial dysfunction.
CLA supplementation among overweight adults significantly reduced body fat over 6 months and prevented weight gain during the holiday season. Although no adverse effects were seen, additional studies should evaluate the effect of prolonged use of CLA.
Recent estimates indicate that 65% of the US adult population is overweight or obese, part of a continuing trend that began in the mid-1980s.1 Overweight and obesity increase all-cause mortality; raise the risk of morbidities including heart disease, stroke, diabetes, osteoarthritis and certain types of cancer; and are primary contributors to preventative death in the US.2 Therefore, determining effective strategies for weight loss or prevention of weight gain is essential.
Conjugated linoleic acid (CLA), an isomer of linoleic acid found in food products of ruminant animals and available as a dietary supplement, was identified as possibly having anticarcinogenic activity by Pariza et al. in 1978.3 Research with the cis-9, trans-11 and trans-10, cis-12 isomers of CLA has shown additional effects in animals, including a decrease in body fat and a reduction in weight gain.4 CLA's effect on body fat was first demonstrated almost a decade ago in mice, and research has since confirmed and extended these findings to include several species, with some variation by animal type and CLA isomer.5 By far, the most dramatic results have been shown in mice with decreases in body fat between 40 and 80% compared to controls.6 Proposed mechanisms have included increased fat mobilization and oxidation, reduced adipocyte proliferation and differentiation, and increased apoptosis in pre-adipocytes.6
Despite clear evidence that CLA reduces body fat in animals, the effect of CLA supplementation on body composition in humans has been inconsistent: of seven studies that have evaluated body composition change with mixed-isomer CLA among overweight or obese adults,7, 8, 9, 10, 11, 12, 13 only three have shown a significant reduction in fat mass.9, 10, 11 The conflicting results in human versus animal studies have been attributed to factors such as lower doses per unit body weight; the use of young, growing animals versus adult humans; and the limited sensitivity and accuracy of in vivo body composition measures.6 In fact, a commonality among previous human studies has been the assessment of body composition with techniques, such as bioelectrical impedance analysis (BIA), dual-energy X-ray absorptiometry (DXA) and underwater weighing, that may be less sensitive in detecting change when compared to criterion multicompartment models.
The aim of the current study was to determine whether CLA supplementation reduces body fat in overweight adults. To address limitations in previous study designs, we utilized the criterion four-compartment model of body composition to assess change in body fat mass. We were also able to determine the effect of CLA on body weight during the holiday season, a period that is commonly associated with weight gain. Research has shown that holiday weight gain contributes considerably to the annual weight gain that may be associated with incremental gains during adulthood, and that overweight and obese individuals are susceptible to greater holiday gains.14
Subjects and methods
Eligible participants were 18- to 44-year-old men and women with body mass index (BMI) values between 25 and 30 kg/m2 at the time of initial screening. Subjects provided medical history and underwent a physical exam, including electrocardiograph and routine blood chemistry analysis. The study was approved by the University of Wisconsin–Madison Institutional Review Board, and all subjects provided written informed consent.
Of the 48 randomized subjects, 40 subjects completed the study protocol and were included in the current data analyses. Eight subjects were excluded: three subjects left the study owing to a lack of interest in participation at months 1, 5 and 6; one subject moved out of the area during month 1; two subjects became ineligible during the course of the study (one started medication to treat depression at month 5 and one resumed smoking during month 1); and two subjects were removed after the completion of the study for non-compliance (one did not meet supplement compliance criteria of 75% and one initiated an intense aerobic training program).
The study design was double-blind, randomized and stratified by gender and age (<30 and 30 years). Subjects were randomly assigned to 4 g/day of placebo (safflower oil) or 4 g/day of 78% active CLA isomers of safflower oil (3.2 g/day CLA: 39.2% cis-9, trans-11 and 38.5% trans-10, cis-12). Each supplement was prepared from the same lot and capsules were identical in color, size and taste (Cognis Corporation, La Grange, IL, USA). Subjects were instructed to take four soft gel capsules each morning with food. At months 0 and 6, subjects arrived in the morning after a 12 h fast and underwent testing for blood chemistry, resting metabolic rate (RMR) and body composition. Body weight, capsule compliance and adverse events (AE) were assessed monthly.
Body composition and body weight
Selinger's four-compartment equation was used to calculate the criterion body fat mass: where %BF is percent body fat, BD is body density (kg/l) determined by underwater weighing, Wt is body mass (kg), TBW is total body water (kg) determined by 18O isotope dilution and TBM is total body mineral (kg) calculated from DXA.15
BD was determined by underwater weighing using procedures described previously.16 Residual lung volume was measured by the closed-circuit oxygen dilution method,17 using a 13.5 l respirometer (Collins Medical, Braintree, MA, USA) and a Med Science nitrogen analyzer (Model 505; St Louis, MO, USA). At month 6, the mean of the month 0 and month 6 residual volumes was used to minimize within-subject variability.
Total Body Water
TBW was measured by 18O dilution using 10.3% 18O-enriched water (Rotem Industries Ltd, Beer Sheva, Israel) and continuous-flow isotope ratio mass spectrometry (Delta S, Finnigan MAT, Bremen, Germany) as described by Schoeller and Luke.18 TBW was calculated from the dilution of the isotopic water using the difference in enrichment between the predose sample and the final (3–4 h) urine sample. Each sample was analyzed on 2 separate days and the average enrichment of the 2 days was used (within-subject standard deviation (s.d.)=0.16 permil)1. The 18O dilution space was assumed to be 0.7% higher than TBW.19
Total Body Mineral
Bone ash was determined by whole-body DXA using the Norland XR-36 bone densitometer with software version 3.7.4/2.1.0 (Norland Corporation, Ft. Atkinson, WI, USA). Because 4.2% of osseous mineral is lost during the ashing process and non-osseous mineral is estimated to be 23.5% of total body bone ash, TBM was calculated by multiplying total body bone ash as determined by DXA by 1.279.20
Height and weight
Height was measured within 0.5 cm on a custom wall-mounted stadiometer and weight within 0.1 kg on a calibrated beam balance platform scale (Continental Scale Corp., Bridgeview, IL, USA).
Abdominal fat mass
DXA was used to assess the change in abdominal fat mass from month 0 to 6 using the Norland XR-36 bone densitometer. The abdominal region was bounded superiorly by the lowest rib and inferiorly by the pubic symphysis.
Energy expenditure and intake
RMR was measured at months 0 and 6 by indirect calorimetry using a ventilated hood (Deltatrac II; SensorMedics Corp., Yorba Linda, CA, USA). RMR was calculated using minute-by-minute O2 and CO2 according to the equation of Weir.21 Physical activity was determined at months 0 and 6 by the 7-day physical activity recall questionnaire according to the methods of Sallis et al.22 Total hours per day of moderate, hard and very hard activity were determined and multiplied by metabolic equivalents (MET) to yield an activity score in MET h/day. Dietary intake was assessed with a 3-day food record at months 0 and 6. Dietary records were analyzed for total energy intake using Food Processor Nutrition Software (version 8.4; ESHA Research, Salem, OR, USA).
Safety and compliance
Blood samples were collected after a 12 h fast on the morning of body composition testing at months 0 and 6. Samples were analyzed for alanine aminotransferase (ALT), alkaline phosphatase (ALP), aspartate aminotransferase (AST), bilirubin, glucose, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, total cholesterol and triglycerides (TGs) (General Medical Laboratories, Madison, WI, USA). Insulin and leptin levels were measured at the University of Wisconsin–Madison Primate Center Assay Laboratory using double-antibody liquid-phase radioimmunoassays. Insulin resistance was determined using the homeostasis model assessment as follows: ((fasting plasma insulin (μIU/ml) × fasting plasma glucose (mmol/l)/22.5).23 Plasma biomarkers of inflammation and endothelial dysfunction were analyzed, including C-reactive protein (CRP), interleukin-6 (IL-6), soluble tumor necrosis factor receptor, E-selectin, soluble intercellular adhesion molecule (sICAM-1) and soluble vascular cell adhesion molecule (sVCAM-1) (Clinical & Epidemiologic Research Laboratory, Boston Children's Hospital).
Adverse events and compliance
AE were recorded monthly throughout the study utilizing a 36-item questionnaire. Symptoms from the questionnaire were then grouped into four categories: cold or flu; general symptoms (nausea, headache); orthopedic symptoms (joint pain, back pain); and emotional symptoms (depression, irritability). Supplement compliance was evaluated monthly by questionnaire and capsule counts. Compliance for inclusion was established a priori as 75%.
The sample size was selected based on 80% power to detect a 1 kg difference in fat mass, with P0.05. The primary outcome variable was the change in body fat mass as measured by the four-compartment model. The t-test procedure was used to determine differences between the groups at month 0 for all variables. Within-group change over 6 months was tested with a paired t-test. Between-group comparisons of the 6-month change values for body composition, blood chemistries, energy expenditure and physical activity were completed using a regression model that adjusted for the covariates of age, sex and baseline BMI. To assess seasonal weight change, weight change was expressed as the average rate per month for the periods of pre-holiday (August–October), holiday (November–December) and post-holiday (January–March). Between-group comparisons in the rate of weight change by season were analyzed with the t-test procedure. Longitudinal weight change by month was analyzed using repeated measures. Differences in the frequency of reported AE were assessed by repeated measures analysis with Poisson regression log-linear model. SAS was used for all analyses (version 8.2; SAS Institute Inc., Cary, NC, USA).
Baseline characteristics, body composition and body weight
There were no baseline differences between the groups in age (CLA, 34±8 years; placebo, 32±7 years) or gender (CLA, five men and 17 women; placebo, three men and 15 women). Body weight, BMI and body composition also did not differ between groups at month 0 (Table 1). The 6-month loss in body fat, expressed either in kilograms (P=0.02) or as a percentage (P=0.02), was significantly greater with CLA as compared to placebo. Within-group comparisons showed significant decreases in fat mass (P=0.05) and body fat percentage (P=0.02) in the CLA group only. There was no difference between the groups in the change in fat-free mass (P=0.8) or abdominal fat mass, although a trend for reduction of abdominal adiposity with CLA was observed (P=0.1). There was a significant difference between groups in the 6-month change in body weight (P=0.04) and BMI (P=0.05). The placebo group had a higher rate of weight gain during the holiday season as compared to the pre-holiday period (P=0.03) and as compared to the CLA group during the holiday season (P=0.01) (Figure 1).
Energy expenditure and intake
There were no differences at month 0 between the groups in RMR, respiratory exchange ratio (RER) or physical activity. Reported energy intake was lower in the placebo group, and this was close to significant (P=0.06). There were no significant differences between the groups in the 6-month change in RMR, RER, reported physical activity or reported total energy intake (Table 2). Within-group changes showed significant increases in RER within the placebo group (P=0.03), and significant increases in RMR within the CLA group (P=0.002). Compared to baseline, reported physical activity decreased in the placebo group by 33% (P=0.07) and in the CLA group by 40% (P=0.001). Both groups also reduced reported energy intake as measured by 3-day food records, with a significant decrease found only within the placebo group (P=0.02).
Safety and compliance
At month 0, there were no significant differences between the groups with one exception: TG levels were higher in the CLA group (Table 3). When two subjects from the CLA group with elevated TGs at month 0 were excluded, the difference between the groups was no longer significant but still showed a tendency toward a difference (P=0.07). Between groups, the only marker that changed significantly was soluble vascular cell adhesion molecule (sVCAM-1), which showed a greater decrease with CLA compared to placebo (P=0.02). Both groups significantly decreased bilirubin, LDL sVCAM-1 cholesterol and total cholesterol, and increased glucose and leptin. Within CLA only, AST decreased and CRP increased. There was one subject within the CLA group with CRP values that increased more than 3 s.d. above the mean, and when this subject was excluded, the within-group increase in CRP was only a trend (P=0.08). In addition, there was no significant difference between the groups in CRP. It is important to note that all within-group changes remained within normal ranges and therefore were not considered clinically significant.
Adverse events and compliance
There was no difference between groups in symptom rate at month 0 (not shown). The frequency of reported AE in the categories of cold or flu, general symptoms and orthopedic symptoms was not significantly different between the groups during the study period. The rate of cold and flu symptoms increased significantly with CLA (P=0.02) and placebo (P<0.0001), which may be a seasonal occurrence, as both groups started in the late summer to early fall and finished in the late winter to early spring. There was a decrease in the frequency of reported emotional symptoms (e.g., anxiety, depression and irritability) in the CLA group that was significantly different from the placebo group (P=0.04). Given the small number of observations, however, it is unlikely that this is a meaningful finding. Six-month compliance rates were 97 and 95% in the placebo group and CLA group, respectively, and there was no significant difference between the groups.
Our results indicate that CLA plays a role in mitigating weight gain associated with the holiday season. CLA significantly reduced body weight compared to placebo during the 6-month study, and during the holiday season in particular. In November and December, the placebo group had a significantly higher rate of weight gain compared to the CLA group and compared to their rate of weight change in the pre-holiday period. A study in the New England Journal of Medicine found that although holiday weight gain averages only 0.4 kg – less than often claimed in the popular press – the weight that is gained during this 6-week period is not reversed during the remainder of the year.14 Thus, holiday weight gain may contribute to the cumulative weight gain that is seen with aging. The study also showed that among overweight and obese individuals, there was a trend toward a large (>2.3 kg) weight gain during the holiday period.14 In our study, CLA versus placebo prevented weight gain during the holiday season among overweight adults and therefore may play a role in reducing the cumulative weight and fat gain that occurs with age.
This is the first study utilizing the criterion four-compartment body composition method to demonstrate changes in body composition due to CLA supplementation in overweight adults. Over 6 months, CLA significantly reduced body fat mass, despite concurrent decreases in reported physical activity. Of seven studies that have examined the effect of mixed-isomer CLA on body composition among overweight or obese subjects,7, 8, 9, 10, 11, 12, 13 three have found similar results.9, 10, 11 Gaullier et al.11 found fat mass losses of 2.4 and 1.7 kg during 1 year of supplementation with CLA in TG and free-fatty acid forms, respectively. In a 1-year, open-label extension of Gaullier's study10 with all three groups taking CLA TG, there was a significant decrease in fat mass in the placebo-TG group (−1.7 kg), whereas the remaining two groups maintained their previous losses and showed additional, but non-significant, decreases in fat mass. During 12 weeks of supplementation, Blankson et al.9 showed that fat mass decreased by 1.2, 1.7 and 1.3 kg for CLA doses of 1.7, 3.4 and 6.8 g/day, respectively.
Conversely, four previous studies of CLA in overweight or obese adults have not shown significant effects on fat mass.7, 8, 12, 13 Berven et al.8 found no significant effect on fat mass with a protocol of 3.4 g/day CLA for 12 weeks, although a trend was observed for body fat to decrease with CLA (0.9 kg) and increase with placebo (0.3 kg). In this study, body composition was assessed with BIA, a method that may be less precise in overweight adults.24 In addition, the study length was possibly too short for an effect to be observed. A commonality among the three remaining studies in overweight adults that did not find an effect of CLA on body fat was the inclusion of a weight loss component: Whigham et al.13 included a 12-week very low-calorie diet (VLCD); Kamphius et al.12 utilized a 3-week VLCD; and Atkinson7 included a diet and exercise regimen during 6 months of treatment. Possibly, the effect of calorie restriction for weight loss in prior studies introduced variability in body weight and body composition change and masked the effect of CLA.
With respect to the safety and tolerance of CLA supplementation, we found no significant effects of CLA on the 6-month change in blood chemistry variables or the frequency of reported AE. Although some animal studies have shown negative effects with CLA, this is possibly attributable to greater doses: animals often receive 10–50 times the dosage per unit body weight compared to humans.25 The effect of CLA on blood glucose and insulin sensitivity is controversial, although reported deleterious effects appear to be primarily related to the trans-10, cis-12 CLA isomer.6 Riserus et al. 26showed detrimental isomer-specific effects on insulin sensitivity in abdominally obese men: supplementation with 3.4 g/day trans-10, cis-12 CLA, but not mixed-isomer CLA, significantly increased fasting glucose and insulin sensitivity as measured by a euglycemic–hyperinsulinemic clamp. Conversely, Eyjolfson27 used an oral glucose tolerance test to demonstrate an improved insulin sensitivity index with 3 g/day mixed-isomer CLA over 8 weeks. Although our study found no effect of mixed-isomer CLA on plasma glucose or insulin levels, a result that has also been supported by several human studies using mixed-isomer supplementation,8, 9, 11, 12, 13 we did not find the expected reductions in glucose or insulin in the CLA-treated group that usually accompanies small weight losses.28 The longest human trial of CLA to date reported no detectable change in glucose and insulin levels during 2 years of supplementation with 3.4 g/day mixed-isomer CLA in the free fatty acid form; yet these individuals failed to show any improvements in glucose and insulin, which might have been expected for their weight loss, and there was a significant increase in insulin within the group taking CLA in the TG form.10 We therefore cannot rule out the potential for small, subclinical effects to develop with prolonged use.
In some animal studies, CLA has altered liver metabolism resulting in increased liver weights and fatty liver.25 We did not see a change in liver function tests including AST, ALT, ALP and total bilirubin. Of six recent studies including similar measures of liver function, only Gaullier et al.10, 11 showed significant within-group increases in AST.8, 9, 10, 11, 13, 29 However, this increase in AST was not significantly different from the control group and was not accompanied by changes in ALT or bilirubin.10, 11 Malpuech et al.29 measured liver size by ultrasound and concluded that CLA had no effect on liver ultrastructure or morphology and no evidence of hepatic lipodystrophy.
CLA has also been shown to reduce cardiovascular disease risk factors such as atherosclerosis and blood lipids in animals, although results have been conflicting.25 In humans, however, the majority of studies have shown that supplementation with CLA mixtures does not have an effect on plasma cholesterol or TG levels compared to controls.5 Similarly, we found no difference between groups in the change in total cholesterol, LDL, HDL or TGs. We also assessed the effect of CLA on biomarkers of inflammation and endothelial dysfunction that have been linked to the development of cardiovascular disease: CRP, IL-6 and TNFR-2 are markers of systemic inflammation, whereas E-selectin, sICAM-1 and s-VCAM-1 are overexpressed in the presence of inflammatory stimuli to the endothelium.30 Although we found a trend for CRP to increase within the CLA group, it is important to note that the CRP value remained within the ‘Average Risk’ category (1–3 mg/l). We also found that s-VCAM-1, which has been described as a predictor of atherosclerotic processes, decreased significantly with CLA.30
The frequency of AE also did not differ between the CLA and placebo groups in our study, although one subject with less than 75% capsule compliance reported difficulty in tolerating the CLA supplement. In the largest human trial of CLA to date (n=157), Gaullier et al.11 also reported no significant difference in the frequency of AE between groups.
There are several limitations to our study design. First, our subjects consisted primarily of overweight women (80%), and although it may not be appropriate to generalize our findings to men, two previous studies found a significant effect of CLA among samples that included a greater percentage of men (33 and 51%).9, 31 Second, we assessed energy intake using a 3-day diet record, and the results of the self-reported energy intakes were counterintuitive. Among the placebo group, the 6-month change in reported energy intake decreased despite a concurrent increase in body weight. In addition, the reported energy intakes among both groups are lower than would be necessary for weight maintenance given the weight and body composition of the subjects.32 It is likely that these self-reported energy intakes are confounded by energy underreporting and should be interpreted with extreme caution.33
In conclusion, we found supplementation with 3.2 g/day of a 50:50 mixture of CLA to be well tolerated and effective in reducing body fat among overweight adults, and we addressed a limitation of previous studies by utilizing the criterion measure of body composition. Further, we found that CLA prevented the seasonal gain of weight that occurred in control subjects, suggesting that CLA may be useful for long-term weight and body composition maintenance by preventing the annual gains that may be especially problematic among overweight individuals. As the longest CLA trial to date is 2 years, however, additional studies are needed before prolonged use of CLA can be recommended.
δ(permil)=((Runknown−Rstandard)/Rstandard)×1000, where R is the ratio of the abundance of heavy isotope to that of the light isotope.
We thank Randall Clark and Jude Sullivan of the University of Wisconsin Hospitals and Clinics Exercise Science Laboratory for conducting body composition analysis; The University of Wisconsin General Clinical Research Center for providing in-patient support and dietary analysis; and Dr Earl Shrago for completing subject medical history and physical examinations. We acknowledge Cognis Nutrition and Health for providing test materials and sponsorship.
ACW contributed to study design, recruited subjects, completed data collection, contributed to data interpretation and wrote the manuscript. ACB contributed to study design and revised the manuscript. RNC assisted with data collection and ZZ provided assistance with statistical analysis. DAS established the study design, obtained study funding, interpreted data and revised the manuscript. None of the authors had or have any financial or personal interest in the sponsor of this research.