Dietary supplementation with cis-9,trans-11 conjugated linoleic acid ...

Dietary supplementation with cis-9,trans-11 conjugated linoleic acid and aortic stiffness in overweight and obese adults1–3 Ivonne Sluijs, Yvonne Plantinga, Baukje de Roos, Louise I Mennen, and Michiel L Bots ABSTRACT Background: Animal studies suggest that dietary cis-9,trans-11 (c9,t11) conjugated linoleic acid (CLA) may inhibit or regress the development of atherosclerosis. The effect of CLA on atherosclerosis has not been assessed in humans. Objective: We investigated the effect of c9,t11 CLA supplementation on aortic pulse wave velocity (a marker of atherosclerosis) and on cardiovascular risk factors in overweight and obese but otherwise apparently healthy subjects. Design: In a double-blind, randomized, placebo-controlled, parallelgroup trial, we randomly assigned 401 subjects, aged 40–70 y and with a body mass index (in kg/m2) 25, to receive either 4 g CLA/d (2.5 g c9,t11 CLA/d and 0.6 g trans-10,cis-12 CLA/d) or placebo supplements for 6 mo. Aortic pulse wave velocity, blood pressure, anthropometric characteristics, and concentrations of fasting lipid, glucose, insulin, and C-reactive protein were measured before and after supplementation. Results: During the intervention, mean (6SE) pulse wave velocity did not change in the c9,t11 CLA group (D0.00 6 0.07) compared with the placebo group (D0.09 6 0.06). There was no effect of c9,t11 CLA supplementation on blood pressure, body composition, insulin resistance, or concentrations of lipid, glucose, and C-reactive protein. Conclusion: This study does not support an antiatherosclerotic effect or an effect on cardiovascular risk factors of c9,t11 CLA. This trial was registered at www.clinicaltrials.gov as NCT00706745. Am J Clin Nutr 2010;91:175–83.

INTRODUCTION

Conjugated linoleic acids (CLAs) are positional and geometric isomers of linoleic acid, in which the 2 double bounds are conjugated (1). CLA is a fatty acid produced in the rumen of ruminants and that occurs naturally in the lipid fraction of ruminant meat and milk. The cis-9,trans-11 (c9,t11) CLA isomer accounts for ’90% of CLA isomers in ruminant fat (2). CLA may have antiatherogenic effects, possibly by acting as an agonist of peroxisome proliferator-activated receptors or as a cyclooxygenase inhibitor, thereby modifying hepatic lipid metabolism and inflammatory activity (3). However, exact mechanisms are unclear. Whereas industrial trans fatty acid consumption has been related to increased risk of cardiovascular disease (CVD), observational studies reported no such associations for ruminant trans fatty acids, such as vaccenic acid and CLA (4). However, this may also be attributed to its low dietary consumption. In-

deed, 2 controlled-feeding trials, which included 38–40 participants, showed a deterioration in blood lipids with daily ruminant trans fatty acid intakes of 3.6–5.0% of energy (5, 6) but not at 1.5% of energy intake (6). Animal studies that investigated CLA mixtures [comprising equal proportions of c9,t11 and trans-10,cis-12 (t10,c12) CLA] at 0.1–5.0% of total dietary intake, consistently related CLA to inhibited atherogenesis (7) and some, but not all, reported improvements in CVD risk markers, such as blood lipids (8). However, divergent effects of the c9,t11 and t10,c12 isomers have been reported, and c9,t11 CLA (at 1.0–2.1%) is the active isomer in the inhibition of atherogenesis and improvement in blood lipid and glucose profiles (9, 10). In humans the c9,t11 isomer in general seems to have no significant effect on CVD risk markers, such as blood lipids (11– 16), inflammatory markers (12–14, 16–18), and adhesion molecules (14, 16, 17, 19). One study reported decreased insulin sensitivity (13), whereas others did not (12, 15, 16, 20). However, because of variations in study design, such as supplementation doses (which varied from 0.6 to 5.5 g/d), the effects of supplementation with a relatively low c9,t11 CLA dose on CVD risk markers are still unclear. Human studies on atherogenesis are scarce (21, 22). One study observed a deterioration of endothelial function, a marker of vascular function, with a mixture of 4.5 g CLA/d (22), but this was not confirmed elsewhere (21). As far as we are aware, no human studies have addressed the effects of CLA, or the c9,t11 isomer, on aortic pulse wave velocity (PWV), which is a measure of aortic stiffness and may be regarded as an early marker of atherosclerosis. Because endothelial dysfunction relates to increased arterial stiffness (23), studies into PWV may add valuable insights into a potential antiatherogenic mechanism of 1 From the Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, Netherlands (IS, YP, and MLB); the University of Aberdeen, Rowett Institute of Nutrition and Health, Aberdeen, United Kingdom (BdR); and Lipid Nutrition, Wormerveer, Netherlands (LIM). 2 Supported by a grant from the Dutch Ministry of Economic Affairs (SENTER-NOVEM, Food and Nutrition Delta Phase 2) and by a grant from Lipid Nutrition BV, Wormerveer, Netherlands. Lipid Nutrition is a commercial company focused on developing and producing specialty fats and oils. 3 Address correspondence to I Sluijs, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Str 6.131, PO Box 85500, 3508 GA Utrecht, Netherlands. E-mail: [email protected]. Received June 8, 2009. Accepted for publication October 23, 2009. First published online November 18, 2009; doi: 10.3945/ajcn.2009.28192.

Am J Clin Nutr 2010;91:175–83. Printed in USA. Ó 2010 American Society for Nutrition

175

176

SLUIJS ET AL

c9,t11 CLA in humans. Therefore, we investigated the effect of 6 mo of supplementation with c9,t11 CLA on aortic PWV. We also studied the effects of c9,t11 CLA on blood pressure, blood lipids, insulin resistance, body composition, and C-reactive protein (CRP) in 346 overweight and obese adults.

SUBJECTS AND METHODS

We conducted a single-center, double-blind, randomized, placebo-controlled, parallel-group trial that compared the effects of c9,t11 CLA supplementation with placebo supplements for 6 mo, on 1) changes in aortic PWV and 2) changes in insulin resistance, blood pressure, anthropometric characteristics, and concentrations of blood lipids, glucose, insulin, and CRP. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki, the International Conference on Harmonization of Good Clinical Practice guidelines, and appropriate regulatory requirements. The study protocol was approved by the Institutional Review Board of the University

Medical Center Utrecht in Utrecht, Netherlands. All participants provided written informed consent. Subjects A total of 1048 potential participants were recruited through the Julius Center "POKA" database (comprising subjects interested in participating in studies) and through a random selection from the municipal registers of a large and middle-large town in the middle part of the Netherlands (Figure 1). Subjects were invited to take part in the study by mail and could reply by a response-card. Interested subjects were contacted by telephone to explain the study aims. Leaflets were mailed with further study information and criteria for eligibility. After 1 wk, these subjects were contacted again and invited to the information and screening/randomization visit. In total, 501 potential participants were screened, of whom 100 were excluded after eligibility criteria were checked, which left 401 subjects to be randomly assigned to study groups (Figure 1). Inclusion criteria at screening were apparently healthy men and women, aged 40–70 y,

FIGURE 1. Disposition of the study population. Values are expressed as n. BP, blood pressure; BPL, blood pressure–lowering; chol, total cholesterol; LL, lipid-lowering; PWV, aortic pulse wave velocity; GL, glucose-lowering; CLA, conjugated linoleic acid; ITT, intention-to-treat.

cis-9,trans-11 CLA AND AORTIC STIFFNESS

with a body mass index (BMI; in kg/m2) of 25. The main exclusion criteria at screening were as follows: a systolic blood pressure of 160 mm Hg or a diastolic blood pressure of 90 mm Hg, or current use of blood pressure–lowering drugs; a total cholesterol concentration of 8 mmol/L or current use of lipidlowering drugs; or inability to perform PWV measurements (Figure 1).

Intervention, compliance, and safety

TABLE 1 Fatty acid composition of placebo and cis-9,trans-11 (c9,t11) conjugated linoleic acid (CLA) oil1

12:0 14:0 16:0 16:1 18:0 18:1 18:2 c9,t11 CLA t10,c12 CLA 1

t10,c12, trans-10,cis-12.

number of capsules that should have been taken in the same period) · 100. Self-reported vital signs and adverse events (AEs) were identified at each clinic visit by means of standard questions. Clinically significant abnormal laboratory values, vital signs, and other physical findings were recorded as AEs. Serious adverse events (SAEs) were monitored continuously throughout the entire study period. PWV measurement

Eligible participants were randomly assigned through a webbased application to receive either c9,t11 CLA or placebo in a 1:1 ratio with the use of a minimalization procedure. Randomization was stratified for sex. Both c9,t11 CLA and placebo supplements were provided in the form of soft gel capsules. The c9,t11 CLA oil was manufactured from safflower oil. Four capsules of 1 g oil each were taken daily. The choice of this dose was not straightforward because extrapolation from animal trials is difficult and no human trials on PWV had been done. We expected to have an effect with a dose of 4 g/d, because this is known to affect other systems in the body, such as fat distribution (8). A lower dose may have had no effect and a higher dose may have decreased compliance. CLA capsules provided 80% CLA isomers, of which 80% (2.5 g/d) was c9,t11 CLA and 20% (0.6 g/d) was t10,c12 CLA. The percentage of energy provided by c9,t11 CLA and t10,c12 CLA during the intervention was ’1% and ’0.3%, respectively (based on a daily intake of 2000–2500 kcal). Because c9,t11 CLA is the abundant isomer in the product, we refer to the product as c9,t11 CLA. The placebo capsules contained an equal amount of fat and were composed of a blend of palm oil (80%) and soybean oil (20%), which resembles the average fatty acid composition of the fat consumed by a Western population (Table 1). Capsules were produced and provided by Lipid Nutrition, Wormerveer, Netherlands. Blinded study capsules were supplied in individual, preprepared, numbered bottles. Subjects were instructed to return leftover trial capsules to the clinic at each visit (at 3 and 6 mo). The investigators checked compliance but were unaware of treatment allocations for the duration of the study. Compliance was calculated with the following formula: compliance (%) = (number of capsules actually taken since last capsule count/

Fatty acid

177

Placebo

CLA

% 0.2 0.8 38.1 0.2 4.0 35.9 18.3 — —

% 0.3 0.2 5.3 0.1 1.1 13.6 2.1 59.4 13.9

PWV was measured at baseline and after 6 mo (or earlier when subjects dropped out of the study) with the use of the SphygmoCor system (PWV Medical, Sydney, Australia), as described by us previously, with well-trained and certified technicians (24–26). Briefly, after the subject rested for 5–10 min in the supine position, PWV was measured by sequential recordings of arterial pressure waveform at the carotid artery and the femoral artery with the use of a hand-held micromanometer-tipped probe on the skin at the site of maximal arterial pulsation. Gating the recordings at those 2 sites to the electrocardiogram allowed PWV to be measured. Recordings were taken when a reproducible signal was obtained with highamplitude excursion (usually 10 consecutive beats) to cover a complete respiratory cycle. The wave transit time was calculated by the system software, with the use of the R wave of a simultaneously recorded electrocardiogram as a reference frame. Distances from the carotid-sampling site to the suprasternal notch and from the suprasternal notch to the femoral artery were measured with a specifically designed compass. The PWV (m/s) was automatically calculated as the distance between the suprasternal notch and the femoral artery minus the distance between the carotid sampling site and the suprasternal notch, divided by the time interval between systolic R wave and femoral systolic up-stroke minus the time interval between systolic R wave and carotid systolic up-stroke. PWV was determined as the mean of 3 consecutive beats recorded during 10 s of data acquisition. All measurements (baseline and end of study) were performed by the same observer. An earlier reproducibility study, performed in 27 participants who underwent 2 PWV measurements within 2 wk (27), showed a mean (6SD) difference of 0.01 6 1.0 m/s between 2 time points and an intraclass correlation coefficient of 89.6%, which indicates that 89.6% of the variance in the PWV measurements was due to patient differences, whereas 10.4% could be attributed to differences between visits. The CV for PWV was 7.5%. Measurement of cardiovascular risk factors At baseline, information on medication use (which included nutritional supplements), medical history (cardiovascular, renal, hepatic, and hematologic diseases), and lifestyle factors (current/ former/nonsmoker and number of cigarettes per day, average number of alcohol units per week) were assessed with the use of a questionnaire. Physical activity scores were calculated with the use of the Modified Baecke Physical Activity Questionnaire, which has been validated in middle-aged and elderly men and women (28, 29). Both questionnaires were filled out by research nurses during a face-to-face interview with the subjects. Blood

178

SLUIJS ET AL

pressure was measured by an oscillometric-automated device (DINAMAP 8100; Critikon, Tampa, FL). Measurements were conducted before 1100 after an overnight fast, performed in duplicate with the participant in a sitting position, with 2 min between each measurement. Systolic and diastolic blood pressures were defined as the average of the 2 measurements. Mean arterial pressure was calculated with the use of the following formula: mean arterial pressure = (2 · mean diastolic blood pressure + mean systolic blood pressure)/3. Pulse pressure was defined as systolic blood pressure minus diastolic blood pressure. Central vascular measurements were performed with the use of the SphygomoCor Blood Pressure Analysis System (Sydney, Australia), as described previously (30). Central aortic pressures were derived by applanation tonometry of the radial artery (31). The radial pressure waveform was recorded noninvasively with a micromanometer (Millar SPT 301; Millar Instruments, Houston, TX). The probe was held on the skin over the maximal arterial pulsation by hand and pressed down on the artery against the underlying bone. Ascending aortic pressure was derived from the central pressure waveform with the use of a generalized transfer function, which is incorporated into the SphygmoCor device (32). The European Systematic Coronary Risk Evaluation (SCORE) formula was used to calculate 10-y absolute risk of fatal CVD for each individual (33). Height and weight were measured once and rounded to the nearest 0.5 cm or 0.5 kg. BMI was calculated as weight in kilograms divided by height in meters squared. Waist circumference was measured at the level of midway between the lower rib and the crista iliaca superior after normal expiration, without pressure of the centimeter at the skin. Hip circumference was measured at the level of the greater trochanter. Both measurements were performed in duplicate and the average taken rounded to the nearest 0.5 cm. Waist-hip ratio was calculated as waist divided by hip circumference in centimeters. After 3 and 6 mo (or earlier when subjects dropped out of the study) mean blood pressure, weight, and waist and hip circumference were measured again. Central vascular measurements, SCORE, and changes in lifestyle (ie, smoking status, physical activity, and alcohol consumption) were recorded at the 6-mo visit. Biochemical analyses Plasma concentrations of total and HDL cholesterol, triglycerides, glucose, and insulin were measured in fasting blood samples at baseline and after 6 mo of intervention. Plasma was obtained by centrifugation (3000 · g, 10 min, 4°C) and stored at 280°C until analysis. An automatic enzymatic procedure was used to determine total and HDL cholesterol (DxC800 Beckman; Beckman Coulter, Mijdrecht, Netherlands). Triglycerides were measured with the use of a homogeneous colorimetric technique (DxC800 Beckman; Beckman Coulter). LDL cholesterol concentrations were estimated with the use of the Friedewald formula (34). We were unable to provide reliable estimates of plasma LDL-cholesterol concentrations for 4 subjects because plasma triglyceride concentrations were .4 mmol/L. High-sensitivity CRP concentrations were determined by an immunoturbidimetric assay on the DxC. Glucose concentrations were assessed capillary with the use of a glucose oxidase method (Accu-Check; Roche, Almere, Netherlands).

Insulin was measured with the use of an immunometric technique on an IMMULITE 1000 Analyzer (Siemens Medical Solutions Diagnostics, Los Angeles, CA). The lower limit of detection was 2 mE/L (13.9 pmol/L), and interassay variation was 9% at 6 mE/L (41.7 pmol/L) and ,5.5% at 20–120 mE/L (138.9– 833.4 pmol/L) (n = 120). We calculated the homeostasis model assessment of insulin resistance (HOMA-IR) to assess insulin resistance, with the equation HOMA-IR = [fasting insulin (mIU/ L) · fasting glucose (mmol/L)]/22.5 (35). Data analysis Earlier studies reported a change in PWV of between 5% and 10% in a period of 6–12 mo on use of statins (36–39). In 60–80-yold men at the Julius Center, we observed a mean (6SD) PWVof 9.0 6 2.5 m/s (26). Because there were no comparable studies on PWV with food supplements, we used these earlier studies on statins to calculate the power. This calculation, with a 2-sided a of 0.05 and power of 80%, showed that ’180 subjects per treatment arm were required to detect an 8.2% difference in change in PWV. The inclusion of a conservatively assumed dropout rate of 15% increased this requirement to ’200 subjects per treatment arm. Continuous baseline characteristics across treatment arms were expressed as mean (6SE) for normally distributed variables and median (range) for non–normally distributed variables. Categoric variables across treatment arms were expressed as n (%). The primary outcome (ie, change in PWV from baseline to 6 mo) was calculated by the subtraction of the 6-mo PWV value from the baseline value. A similar approach was taken for the secondary outcomes. Changes from baseline to 6 mo across treatment arms were expressed as mean (6SE) or n (%) and analyzed in accordance with the intention-to-treat principle. Differences in 6-mo changes between the 2 groups were tested with the use of a 2-sided t test or chi-square test. Within-group changes were tested with a paired t test, Wilcoxon’s test, or chi-square test. Because of randomized allocation, no adjustments for baseline levels were made. The analyses were repeated among compliant subjects only and with the use of the per-protocol principle. Furthermore, analyses were repeated with imputed values for missing follow-up PWV values (n = 48) because missing data may not occur at random and may bias the final results (40). Imputation was performed with the use of linear regression modeling. Statistical analyses were done with the use of SPSS for Windows, version 14.0 (SPSS Inc, Chicago, IL). RESULTS

In total, 401 subjects were randomly assigned: 200 to the placebo group and 201 to the c9,t11 CLA group. After they were randomly assigned, 7 subjects (4 in the placebo group and 3 in the c9,t11 CLA group) were excluded because they met the exclusion criteria and should therefore not have been assigned to a group. After 6 mo of follow-up, 45 subjects (11%) had discontinued consumption of the capsules. Discontinuation was similar in both groups. Furthermore, 48 subjects (23 in the placebo group and 25 in the c9,t11 CLA group) had missing 6-mo follow-up PWV measurements. Finally, 346 subjects were included in the analyses, 173 in the control group and 173 in the c9,t11 CLA group (Figure 1). The mean compliance rate was

179


93% in the placebo group and 92% in the c9,t11 CLA group. Both arms had similar mean (6SE) follow-up times between baseline and postbaseline PWV measurements (control group: 179 6 2 d; c9,t11 CLA group: 180 6 2 d). Demographic characteristics and CVD risk markers at baseline were similar between the 2 groups (Table 2). After 6 mo of supplementation, mean (6SE) PWV did not change in the c9,t11 CLA group compared with the placebo group (D0.00 6 0.07 compared with D0.09 6 0.06 m/s) (Table 3). Results did not differ when analyses were performed in accordance with the per-protocol principle (n = 172 in each group) among compliant subjects only (n = 146 in the placebo group and 141 in the c9,t11 CLA group) or after imputation of missing values (data not shown). We observed no interaction with sex, SCORE, degree of obesity, or pretest PWV. Pearson correlation coefficients between changes in PWV and changes in blood pressure, mean arterial pressure, pulse pressure, and central vascular measures varied from 0.05 to 0.26 in the c9,t11 CLA group and from 0.01 to 0.25 in the placebo group. There were no significant differences in changes in mean systolic and diastolic blood pressure, pulse pressure, and mean arterial pressure between the c9,t11 CLA and placebo groups; neither were there between-group differences in central aortic systolic and diastolic blood pressure or central aortic pulse pressure (Table 3). The effects of 6 mo of supplementation on

body weight, BMI, waist circumference, and waist-hip ratio did not differ significantly between the c9,t11 CLA and placebo groups. However, we observed an increase in all of these measures in the placebo group, and in waist circumference in the c9,t11 CLA group (Table 3). Plasma concentrations of triglycerides, total, LDL, and HDL cholesterol, and total/HDL cholesterol ratio did not change significantly in the c9,t11 CLA group compared with the placebo group. There were no significant between-group differences in changes in plasma glucose, insulin, HOMA-IR, or CRP (Table 3). When individuals with CRP concentrations .10 mg/L were excluded, the findings for CRP did not change. In total, 188 AEs were reported by 102 subjects of the c9,t11 CLA group and 170 AEs by 94 subjects in the placebo group, without a statistically significant difference (P = 0.59). Seven SAEs were reported in the c9,t11 CLA arm and 8 SAEs in the placebo group. All SAEs were judged to be unrelated to the use of c9,t11 CLA. DISCUSSION

In this 6-mo, double-blind, randomized, placebo-controlled, parallel-group trial, we observed no significant effect of c9,t11 CLA supplementation on aortic PWV, blood pressure, plasma lipids, glucose, insulin resistance, body composition, or CRP in apparently healthy but overweight and obese adults.

TABLE 2 Baseline characteristics of the study population by treatment assignment (intention-to-treat)1

Age (y) Men [n (%)] Aortic pulse wave velocity (m/s) Central aortic systolic pressure (mm Hg) Central aortic diastolic pressure (mm Hg) Central aortic pulse pressure (mm Hg) Mean systolic pressure (mm Hg) Mean diastolic pressure (mm Hg) Mean pulse pressure (mm Hg) Mean arterial pressure (mm Hg) Weight (kg) BMI (kg/m2) Waist circumference (cm) Waist-hip ratio Plasma total cholesterol (mmol/L) Plasma LDL cholesterol (mmol/L) Plasma HDL cholesterol (mmol/L) Plasma total/HDL-cholesterol ratio (mmol/L) Plasma fasting triglycerides (mmol/L) Plasma fasting glucose (mmol/L) Plasma fasting insulin (pmol/L) HOMA-IR hs-CRP (mg/L)4 Current smoking [n (%)] Baecke physical activity score4 SCORE 1

Placebo (n = 173)

c9,t11 CLA (n = 173)

58.8 6 0.5 84 (48.6) 7.80 6 0.11 126.8 6 1.2 79.5 6 0.7 47.3 6 0.9 126.8 6 1.0 75.7 6 0.6 51.1 6 0.8 92.7 6 0.7 85.2 6 1.0 27.7 (24.7, 50.2)3 99.2 6 0.7 0.94 6 0.01 5.66 6 0.07 3.84 6 0.06 1.28 6 0.03 4.69 6 0.09 1.00 (0.30, 3.70) 5.26 6 0.04 62.5 (13.9, 236.1) 2.02 (0.41, 9.52) 1.68 (0.15, 25.10) 14 (8.1) 7.9 (2.8, 10.8) 2 (0, 12)

58.0 6 0.4 83 (48.0) 7.79 6 0.10 126.6 6 1.0 79.9 6 0.6 46.7 6 0.7 128.0 6 1.0 76.5 6 0.6 51.5 6 0.7 93.7 6 0.7 85.6 6 0.9 28.0 (24.9, 43.8) 99.0 6 0.7 0.94 6 0.01 5.47 6 0.06 3.68 6 0.06 1.29 6 0.03 4.47 6 0.09 1.00 (0.40, 4.60) 5.15 6 0.04 55.6 (13.9, 361.1) 1.84 (0.44, 10.86) 1.44 (0.15, 38.20) 18 (10.4) 7.8 (4.0, 10.4) 2 (0, 12)

2

c9,t11, cis-9,trans-11; CLA, conjugated linoleic acid; HOMA-IR, homeostasis model assessment of insulin resistance; hs-CRP, high-sensitivity C-reactive protein; SCORE, Systematic Coronary Risk Evaluation. 2 Mean 6 SD (all such values). 3 Median; range in parentheses (all such values). 4 n = 168 in placebo group.

180

SLUIJS ET AL

TABLE 3 Changes in pulse wave velocity (PWV) and cardiovascular risk markers from baseline to 6 mo by treatment assignment and between-group differences (intention-to-treat)1

Aortic PWV (m/s) Central aortic systolic pressure (mm Hg) Central aortic diastolic pressure (mm Hg) Central aortic pulse pressure (mm Hg) Mean systolic pressure (mm Hg) Mean diastolic pressure (mm Hg) Mean pulse pressure (mm Hg) Mean arterial pressure (mm Hg) Weight (kg) BMI (kg/m2) Waist circumference (cm) Waist-hip ratio Plasma total cholesterol (mmol/L) Plasma LDL cholesterol (mmol/L)6 Plasma HDL cholesterol (mmol/L) Plasma total/HDL-cholesterol ratio (mmol/L) Plasma fasting triglycerides (mmol/L) Plasma fasting glucose (mmol/L) Plasma fasting insulin (pmol/L)7 HOMA-IR8 hs-CRP (mg/L)8 Current smoking [n (%)] Baecke physical activity score9 SCORE

Placebo (n = 173)

c9,t11 CLA (n = 173)

0.09 6 0.063 20.3 6 0.8 20.3 6 0.5 0.0 6 0.7 20.30 6 0.85 20.25 6 0.48 20.05 6 0.66 20.27 6 0.55 0.65 6 0.165 0.22 6 0.055 0.79 6 0.255 0.005 6 0.0025 20.06 6 0.06 20.03 6 0.05 20.04 6 0.025 0.08 6 0.06 0.04 6 0.04 20.12 6 0.045 1.83 6 1.96 0.03 6 0.07 0.06 6 0.52 22 (1.2) 20.13 6 0.065 0.19 6 0.085

0.00 6 0.07 0.2 6 0.8 0.7 6 0.5 20.5 6 0.7 21.04 6 0.76 20.21 6 0.50 20.83 6 0.63 20.49 6 0.52 0.21 6 0.22 0.08 6 0.08 0.84 6 0.315 0.005 6 0.003 0.02 6 0.06 0.00 6 0.05 20.02 6 0.02 0.13 6 0.05 0.08 6 0.035 20.03 6 0.04 0.20 6 2.31 0.01 6 0.08 0.32 6 0.42 21 (0.6) 20.12 6 0.06 0.09 6 0.09

Between-group differences 20.09 0.5 1 20.5 20.74 0.04 20.78 20.22 20.44 20.14 0.05 0.0 0.08 0.03 0.02 0.05 0.04 0.09 21.63 20.02 0.26 1 0.01 20.10

(20.29, 0.09)4 (21.8, 2.8) (20.4, 2.4) (22.5, 1.5) (22.98, 1.50) (21.33, 1.40) (22.58, 1.02) (21.70, 1.26) (20.99, 0.10) (20.32, 0.04) (20.73, 0.84) (20.007, 0.008) (20.08, 0.25) (20.10, 0.16) (20.03, 0.07) (20.10, 0.19) (20.07, 0.14) (20.02, 0.20) (27.59, 4.34) (20.23, 0.18) (21.05, 1.57) (0.6) (20.15, 0.17) (20.31, 0.11)

P value2 0.31 0.68 0.15 0.61 0.51 0.96 0.40 0.77 0.11 0.14 0.89 0.91 0.34 0.62 0.41 0.56 0.55 0.10 0.59 0.82 0.69 0.56 0.88 0.35

1

c9,t11, cis-9,trans-11; CLA, conjugated linoleic acid; HOMA-IR, homeostasis model assessment of insulin resistance; hs-CRP, high-sensitivity Creactive protein; SCORE, Systematic Coronary Risk Evaluation. 2 For differences between the c9,t11 CLA and control group, P values were obtained by using a 2-sided t test for means 6 SEs and a chi-square test for n (%). 3 Mean 6 SE (all such values). 4 Mean; 95% CI in parentheses (all such values). 5 Significant within-group change from baseline, P , 0.05 (paired t test). 6–9 For placebo and c9,t11 CLA groups, respectively: 6n = 170 and n = 172, 7n = 167 and n = 170, 8n = 168 and n = 170, 9n = 166 and n = 166.

The strengths of our study include its large sample size and its double-blind, randomized, placebo-controlled design, which decreased the possibility of bias. However, some aspects of the study need to be addressed. First, we had no information on dietary intake at baseline and changes in diet during the intervention. Although we cannot exclude differences in baseline dietary intake between the 2 groups, we expect intake to be similar because of the large sample size and randomized design of our study. Furthermore, although participants were requested not to change dietary habits during the intervention period, we cannot exclude the possibility that between-group differences in dietary intake during the intervention may have influenced our results. This would be particularly relevant if the 2 interventions had been calorically different from each other. Because this is not the case, we assume that potential changes in dietary intake were equal in both groups. Second, in our study we focused on the effect of c9,t11 CLA by using an investigational product that consisted of 2.5 g c9,t11 CLA and 0.6 g t10,c12 CLA. Although the t10,c12 CLA content of the product was relatively low and probably too low for any of the effects to be attributed to it (15), we cannot exclude the possibility that our results may be partly attributable to this isomer as well. This study is, to our knowledge, the first to investigate the effect of supplementation with a predominantly single isomer,

c9,t11 CLA, on aortic PWV, and we observed no significant effect after 6 mo. Raff et al (21) reported that consumption of a CLA (mixture)-rich diet for 5 wk, which provided 4.7 g of c9,t11 CLA and t10,c12 CLA isomers in equal amounts per day, did not affect brachial endothelial function, as another early marker of atherosclerosis, in 60 young and healthy men, which agrees with our finding. However, Taylor et al (22) observed that endothelial function, assessed by brachial artery flow-mediated dilatation, worsened after 12 wk of supplementation with a mixture of 4.5 g CLA/d, comprising approximately equal proportions of c9,t11 CLA and t10,c12 CLA, in 40 middle-aged overweight and obese men. Potential explanations for this difference from our findings may be differences in proportion of CLA isomers provided [50:50 compared with 80:20 proportion of c9,t11 and t10,c12 CLA (9, 10)] and in study endpoints. Because endothelial dysfunction is a process that occurs very early on in the development of atherosclerosis (41) as compared with arterial stiffness, it might be that endothelial function has relation with our intervention. Several reasons for no detection of a statistically significant effect of c9,t11 CLA on PWV need to be considered before we can exclude an effect from this study. First, the power is mainly driven by the anticipated treatment effect and by the rate of change. The power calculation for our study was based on the


effects of statin use for 6 mo to 1 y on PWV (36–39), which suggests that we would be able to detect an 8% difference between 2 groups of 180 subjects each. However, the effect produced by CLA supplementation was more subtle and maybe more in line with what could be expected from a food supplement. However, the observed change in aortic PWV in the placebo group in a 6-mo period was comparable to that which has been reported elsewhere (36–39) or which could be extrapolated from cross-sectional studies that looked into aging effects (27, 42). To evaluate an observed 1.2% difference between the groups, 1264 subjects per group would have been needed. This clearly reflects the fact that CLA is unlikely to affect PWV. Unfortunately, no comparable studies that investigated the effect of nutritional intervention on PWV have been performed so far. Second, the dose of CLA provided in our study (2.5 g c9,t11 CLA) may have been too low to induce significant effects. The antiatherogenic effects of c9,t11 CLA in animal studies were based on dietary doses that ranged from 0.5% to 2.1% of energy (9, 10, 43–47), which would be equivalent to supplementation doses of 1.4–5.8 g CLA in humans (based on a daily intake of 2500 kcal). Third, it is unlikely that our results are influenced by subjects who discontinued taking study capsules (1%) or by dropouts caused by missing follow-up PWV values (12%), because analyses carried out in accordance with the per-protocol principle and with imputed values for missing PWV values revealed similar results. Attenuation of observed effects due to noncompliance (10%) is not likely because analyses restricted to compliant subjects showed results comparable to the intention-to-treat analyses. Fourth, although our study included overweight and obese participants, the absence of an effect may also be attributed to our participants being generally healthy, with relatively few smokers, and with normal plasma lipid and glucose concentrations. Animal studies that reported effects of CLA on atherosclerosis were mainly performed in animals models prone to atherosclerosis development (8). It is important to note that changes in PWV may be (partly) a result of changes in its determinants, such as blood pressure and body weight. We observed no effect of c9,t11 CLA on these determinants, which may also explain the lack of a statistically significant effect on aortic PWV. Others have also reported no effect of a CLA mixture on systolic and diastolic blood pressure (21, 48). It is known that aortic PWV and atherosclerosis can be affected via other pathways, such as inflammation (49), which indicates that c9,t11 CLA may still have an effect on atherosclerosis through this pathway. However, we did not find an effect of c9,t11 CLA on CRP. Other human trials have not observed an effect of 1.4–5.1 g c9,t11 CLA/d (14, 16, 18). Thus far, most human intervention studies have suggested little evidence for an effect of c9,t11 CLA on CVD risk markers, such as plasma lipid concentrations, body composition, and insulin resistance (8, 11–16, 20, 50), although one study, which included 25 abdominally obese men, observed a decrease in insulin sensitivity after 3 mo of supplementation with 3.2 g c9,t11 CLA/d (13). However, studies performed thus far have been relatively small in number (25–90 subjects), with large variations in study duration (3–18 wk), intervention doses of c9,t11 CLA (0.6–5.5 g/d), types of placebo, and subject characteristics such as sex, age, and BMI range. Our study, with a substantially larger study population of persons who

181

were either overweight or obese, and with a relatively long study duration, however, confirmed the neutral effects of c9,t11 CLA on blood lipid concentrations, body composition, and insulin resistance. CLA can be considered a ruminant-derived trans fatty acid. Observational studies suggested no association between CVD risk and ruminant trans fatty acids but increased risk with industrial trans fatty acids (4). However, 2 recent dietary-controlled trials that compared ruminant and industrial trans fatty acids reported deterioration of blood lipids with high doses (3.6–5.0% of energy) of both trans fatty acid sources (5, 6) but not with a moderate dose (1.5% of energy) of ruminant trans fatty acids (6), which suggests that dose is more important than source. Although we did not address industrial trans fatty acids, our study supports the lack of effect on CVD risk markers with relatively low doses of ruminant trans fatty acids. The dose administered, 2.5 g c9,t11 CLA/d and 0.6 g t10,c12 CLA/d, might be considered a safe dose with regard to CVD risk. CLA concentrations in ruminant meat and milk are ,1% of total fat (51). Consequently, human CLA intake is quite low [212 mg/d in men and 151 mg/d in women (52)], and it would be difficult to achieve CLA intakes comparable to the dose investigated in this study. However, an alteration in ruminant feeding could be used to increase CLA intakes (53). Nevertheless, because we observed no effects on PWV or other CVD risk markers, our study does not provide evidence for a role of c9,t11 CLA in CVD risk reduction. In conclusion, this study does not support an antiatherosclerotic effect or an effect on other CVD risk markers of c9,t11 CLA in overweight and obese adults. We thank the personnel of the Clinical Trial Unit of the Julius Center for their excellent contribution to the conduct of the trial and the data collection. The authors’ responsibilities were as follows—YP, BdR, LIM, and MLB: study concept and design; YP: provision of training in how to take PWV measurements; YP and MLB: data acquisition; IS: statistical analyses; IS, YP, and MLB: interpretation of results; IS and MLB: draft of manuscript; BdR, LIM, and MLB: critical review of manuscript; and MLB: study supervision. IS, YP, and BdR had no conflict of interest. LIM was employed by Lipid Nutrition at the start of the study and currently runs a nutrition consultancy agency. MLB runs a core laboratory on measurements of arterial stiffness in observational and intervention studies; apart from receiving the grant for performing this study, MLB had no other conflicts to disclose. Lipid Nutrition was involved in the initiation of the study. The sponsor was given the drafts of the manuscript for comments and suggestions, but the final decision was at the discretion of the main principal investigator (MLB).

REFERENCES 1. Banni S. Conjugated linoleic acid metabolism. Curr Opin Lipidol 2002; 13:261–6. 2. Chin SF, Liu W, Storkson JM, Ha YL, Pariza MW. Dietary sources of conjugated dienoic isomers of linoleic acid, a newly recognized class of anticarcinogens. J Food Compos Anal 1992;5:185–97. 3. Belury MA. Dietary conjugated linoleic acid in health: physiological effects and mechanisms of action. Annu Rev Nutr 2002;22:505–31. 4. Mozaffarian D, Aro A, Willett WC. Health effects of trans-fatty acids: experimental and observational evidence. Eur J Clin Nutr 2009;63(suppl 2):S5–21. 5. Chardigny JM, Destaillats F, Malpuech-Brugere C, et al. Do trans fatty acids from industrially produced sources and from natural sources have the same effect on cardiovascular disease risk factors in healthy

182

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

SLUIJS ET AL

subjects? Results of the trans Fatty Acids Collaboration (TRANSFACT) study. Am J Clin Nutr 2008;87:558–66. Motard-Belanger A, Charest A, Grenier G, et al. Study of the effect of trans fatty acids from ruminants on blood lipids and other risk factors for cardiovascular disease. Am J Clin Nutr 2008;87:593–9. Mitchell PL, McLeod RS. Conjugated linoleic acid and atherosclerosis: studies in animal models. Biochem Cell Biol 2008;86:293–301. Bhattacharya A, Banu J, Rahman M, Causey J, Fernandes G. Biological effects of conjugated linoleic acids in health and disease. J Nutr Biochem 2006;17:789–810. Arbones-Mainar JM, Navarro MA, Guzman MA, et al. Selective effect of conjugated linoleic acid isomers on atherosclerotic lesion development in apolipoprotein E knockout mice. Atherosclerosis 2006; 189:318–27. de Roos B, Rucklidge G, Reid M, et al. Divergent mechanisms of cis9, trans11-and trans10, cis12-conjugated linoleic acid affecting insulin resistance and inflammation in apolipoprotein E knockout mice: a proteomics approach. FASEB J 2005;19:1746–8. Malpuech-Brugere C, Verboeket-van de Venne WP, Mensink RP, et al. Effects of two conjugated linoleic acid isomers on body fat mass in overweight humans. Obes Res 2004;12:591–8. Naumann E, Carpentier YA, Saebo A, et al. cis-9, trans-11 and trans-10, cis-12 conjugated linoleic acid (CLA) do not affect the plasma lipoprotein profile in moderately overweight subjects with LDL phenotype B. Atherosclerosis 2006;188:167–74. Riserus U, Vessby B, Arnlov J, Basu S. Effects of cis-9,trans-11 conjugated linoleic acid supplementation on insulin sensitivity, lipid peroxidation, and proinflammatory markers in obese men. Am J Clin Nutr 2004;80:279–83. Tholstrup T, Raff M, Straarup EM, Lund P, Basu S, Bruun JM. An oil mixture with trans-10, cis-12 conjugated linoleic acid increases markers of inflammation and in vivo lipid peroxidation compared with cis-9, trans-11 conjugated linoleic acid in postmenopausal women. J Nutr 2008;138:1445–51. Tricon S, Burdge GC, Kew S, et al. Opposing effects of cis-9,trans-11 and trans-10,cis-12 conjugated linoleic acid on blood lipids in healthy humans. Am J Clin Nutr 2004;80:614–20. Tricon S, Burdge GC, Jones EL, et al. Effects of dairy products naturally enriched with cis-9,trans-11 conjugated linoleic acid on the blood lipid profile in healthy middle-aged men. Am J Clin Nutr 2006;83:744–53. Nugent AP, Roche HM, Noone EJ, Long A, Kelleher DK, Gibney MJ. The effects of conjugated linoleic acid supplementation on immune function in healthy volunteers. Eur J Clin Nutr 2005;59:742–50. Ramakers JD, Plat J, Sebedio JL, Mensink RP. Effects of the individual isomers cis-9,trans-11 vs. trans-10,cis-12 of conjugated linoleic acid (CLA) on inflammation parameters in moderately overweight subjects with LDL-phenotype B. Lipids 2005;40:909–18. Watras AC, Buchholz AC, Close RN, Zhang Z, Schoeller DA. The role of conjugated linoleic acid in reducing body fat and preventing holiday weight gain. Int J Obes 2007;31:481–7. Noone EJ, Roche HM, Nugent AP, Gibney MJ. The effect of dietary supplementation using isomeric blends of conjugated linoleic acid on lipid metabolism in healthy human subjects. Br J Nutr 2002;88:243–51. Raff M, Tholstrup T, Sejrsen K, Straarup EM, Wiinberg N. Diets rich in conjugated linoleic acid and vaccenic acid have no effect on blood pressure and isobaric arterial elasticity in healthy young men. J Nutr 2006;136:992–7. Taylor JS, Williams SR, Rhys R, James P, Frenneaux MP. Conjugated linoleic acid impairs endothelial function. Arterioscler Thromb Vasc Biol 2006;26:307–12. McEniery CM, Wallace S, Mackenzie IS, et al. Endothelial function is associated with pulse pressure, pulse wave velocity, and augmentation index in healthy humans. Hypertension 2006;48:602–8. van den Elzen AP, Sierksma A, Oren A, et al. Alcohol intake and aortic stiffness in young men and women. J Hypertens 2005;23:731–5. van der Schouw YT, Pijpe A, Lebrun CE, et al. Higher usual dietary intake of phytoestrogens is associated with lower aortic stiffness in postmenopausal women. Arterioscler Thromb Vasc Biol 2002;22: 1316–22. Nakhai-Pour HR, Grobbee DE, Bots ML, Muller M, van der Schouw YT. C-reactive protein and aortic stiffness and wave reflection in middle-aged and elderly men from the community. J Hum Hypertens 2007;21:949–55.

27. Lebrun CE, van der Schouw YT, Bak AA, et al. Arterial stiffness in postmenopausal women: determinants of pulse wave velocity. J Hypertens 2002;20:2165–72. 28. Pols MA, Peeters PH, Bueno-De-Mesquita HB, et al. Validity and repeatability of a modified Baecke questionnaire on physical activity. Int J Epidemiol 1995;24:381–8. 29. Hertogh EM, Monninkhof EM, Schouten EG, Peeters PH, Schuit AJ. Validity of the Modified Baecke Questionnaire: comparison with energy expenditure according to the doubly labeled water method. Int J Behav Nutr Phys Act 2008;5:30. 30. Nijdam ME, Plantinga Y, Hulsen HT, et al. Pulse pressure amplification and risk of cardiovascular disease. Am J Hypertens 2008;21:388–92. 31. O’Rourke MF, Gallagher DE. Pulse wave analysis. J Hypertens Suppl 1996;14:S147–57. 32. Chen CH, Nevo E, Fetics B, et al. Estimation of central aortic pressure waveform by mathematical transformation of radial tonometry pressure: validation of generalized transfer function. Circulation 1997;95: 1827–36. 33. Conroy RM, Pyorala K, Fitzgerald AP, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003;24:987–1003. 34. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502. 35. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–9. 36. Maki-Petaja KM, Booth AD, Hall FC, et al. Ezetimibe and simvastatin reduce inflammation, disease activity, and aortic stiffness and improve endothelial function in rheumatoid arthritis. J Am Coll Cardiol 2007;50: 852–8. 37. Ichihara A, Hayashi M, Koura Y, Tada Y, Kaneshiro Y, Saruta T. Long-term effects of statins on arterial pressure and stiffness of hypertensives. J Hum Hypertens 2005;19:103–9. 38. Pirro M, Schillaci G, Mannarino MR, et al. Effects of rosuvastatin on 3-nitrotyrosine and aortic stiffness in hypercholesterolemia. Nutr Metab Cardiovasc Dis 2007;17:436–41. 39. Raison J, Rudnichi A, Safar ME. Effects of atorvastatin on aortic pulse wave velocity in patients with hypertension and hypercholesterolaemia: a preliminary study. J Hum Hypertens 2002;16:705–10. 40. Little RA, Rubin DB. Statistical analysis with missing data. New York, NY: Wiley, 1987. 41. Moreno PR, Sanz J, Fuster V. Promoting mechanisms of vascular health: circulating progenitor cells, angiogenesis, and reverse cholesterol transport. J Am Coll Cardiol 2009;53:2315–23. 42. McEniery CM. Yasmin, Hall IR, Qasem A, Wilkinson IB, Cockcroft JR. Normal vascular aging: differential effects on wave reflection and aortic pulse wave velocity: the Anglo-Cardiff Collaborative Trial (ACCT). J Am Coll Cardiol 2005;46:1753–60. 43. Kritchevsky D, Tepper SA, Wright S, Czarnecki SK, Wilson TA, Nicolosi RJ. Conjugated linoleic acid isomer effects in atherosclerosis: growth and regression of lesions. Lipids 2004;39:611–6. 44. Nestel P, Fujii A, Allen T. The cis-9,trans-11 isomer of conjugated linoleic acid (CLA) lowers plasma triglyceride and raises HDL cholesterol concentrations but does not suppress aortic atherosclerosis in diabetic apoE-deficient mice. Atherosclerosis 2006;189:282–7. 45. Toomey S, Harhen B, Roche HM, Fitzgerald D, Belton O. Profound resolution of early atherosclerosis with conjugated linoleic acid. Atherosclerosis 2006;187:40–9. 46. Valeille K, Ferezou J, Amsler G, et al. A cis-9,trans-11-conjugated linoleic acid-rich oil reduces the outcome of atherogenic process in hyperlipidemic hamster. Am J Physiol Heart Circ Physiol 2005;289: H652–9. 47. Valeille K, Ferezou J, Parquet M, et al. The natural concentration of the conjugated linoleic acid, cis-9,trans-11, in milk fat has antiatherogenic effects in hyperlipidemic hamsters. J Nutr 2006;136:1305–10. 48. Iwata T, Kamegai T, Yamauchi-Sato Y, et al. Safety of dietary conjugated linoleic acid (CLA) in a 12-weeks trial in healthy overweight Japanese male volunteers. J Oleo Sci 2007;56:517–25. 49. Yasmin, McEniery CM, Wallace S, Mackenzie IS, Cockcroft JR, Wilkinson IB. C-reactive protein is associated with arterial stiffness in apparently healthy individuals. Arterioscler Thromb Vasc Biol 2004;24: 969–74.

cis-9,trans-11 CLA AND AORTIC STIFFNESS 50. Mensink RP. Metabolic and health effects of isomeric fatty acids. Curr Opin Lipidol 2005;16:27–30. 51. Dhiman TR, Nam SH, Ure AL. Factors affecting conjugated linoleic acid content in milk and meat. Crit Rev Food Sci Nutr 2005;45:463–82. 52. Ritzenthaler KL, McGuire MK, Falen R, Shultz TD, Dasgupta N, McGuire MA. Estimation of conjugated linoleic acid intake by written

183

dietary assessment methodologies underestimates actual intake evaluated by food duplicate methodology. J Nutr 2001;131:1548–54. 53. Duckett SK, Andrae JG, Owens FN. Effect of high-oil corn or added corn oil on ruminal biohydrogenation of fatty acids and conjugated linoleic acid formation in beef steers fed finishing diets. J Anim Sci 2002; 80:3353–60.