Evaluation of Health Benefits of Conjugated Linoleic Acid (CLA)-Enriched Dairy and Beef Foods
NE Iowa RC&D, Inc., Iowa State University, Coolee Region Organic Producers, and local producers joined to carry out a second project that will be a human feeding study using CLA enriched food items verses non-CLA items. The purpose of this research will be to verify the health benefits of this acid in humans, to quantify how much CLA needs to be ingested to have an effect on health, as well as creating new practices for farmers to enhance the diets of their livestock to obtain an increased level of CLA in the final product.
This project will conduct an experiment that evaluates CLA-enriched dairy and beef foods from farmers on contract with CROPP. The farmers will use grazing systems to raise livestock and produce dairy products with high CLA content for the experiment. The milk and beef they produce will be processed and then prepared to create several different foods that will be incorporated into typical U.S. diets for a human feeding study. The study then will examine the health benefits of humans consuming CLA-enriched dairy and beef foods by evaluating the effect of feeding diets containing these CLA-enriched foods on concentration of cholesterol, individual lipoproteins, glucose, glucagon, and insulin in blood plasma, bone density, body weight and composition, and glucose tolerance. Results will be widely distributed through publication, field days and presentations. Several positive project outcomes are expected.
Short-term outcomes of the project will enhance the quality of life for participants and encourage sustainable agricultural systems. Specific short-term outcomes include:
Improved human health, quantified through the direct effects of CLA-enriched foods on measures of human health.
Because pasture feeding is the management system found to produce CLA-enriched foods, (outcome of our current SARE-funded project), intermediate outcomes of this research will improve the profitability for farmers that utilize sustainable management systems including:
Greater economic viability for dairy and beef producers that use sustainable high CLA pasture systems. Once the benefit to humans is demonstrated, high CLA content meat and dairy products are expected to increase in market value. Therefore, producers using the sustainable, high CLA pasture system, will be able to obtain a premium price in new and emerging markets and increase the market share for high CLA content pasture-fed animals.
The long-term outcomes of this project will enhance the quality of life for farmers and ranchers, rural communities, and society as a whole by providing a better understanding of CLA as an indicator that links healthy lands and healthy animals to healthy humans. Long-term benefits will include:
Improved health of the land and of human communities. This will be accomplished through a demonstration of the economic, environmental, and health justifications of management systems that result in greater utilization of pasture-based feeding systems for production of animal-derived foods on marginal land.
Spring 2007 –
Over 10,000 women associated with Iowa State University were contacted via email, fliers, and word-of-mouth encouraging them to pursue being screened for the study. A number of women were excluded on the basis of age (outside of 20-40 years of age), body mass index, eating habits/preferences, schedules, and other parameters that conflicted with the study. Blood samples were collected from 56 women to screen for total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. These samples were sent to LabCorp for clinical analysis. Of the 56 women, 22 showed increased total cholesterol and increased LDL cholesterol concentrations. These 22 women were asked to, and agreed to, participate in the study.
Beef – Ground beef from pasture-fed and grain-fed steers had been previously procured and stored at -20ºC. This meat, which had been used in the first study, was deemed safe, unaltered, and edible for the second study. Concern arose when packages of pasture-fed ground beef were shown to contain only 8% lipid on a wet basis. The boxes of meat from which these packages came were set aside and not used for the study, using only boxes from which the packages of meat contained 15% lipid. The latter matched the grain-fed beef.
Cheese – Cheese from a pasture-fed dairy was procured and stored at -20ºC from the previous study, and this cheese, like the beef, was found to be safe, unaltered, and edible for the second study. Several commercial cheeses were analyzed to determine which commercial brand contains the most similar macronutrient composition to the pasture-fed cheese. It was found that FastCo cheddar cheese was most similar, and was therefore used on the study.
Ice Cream – Haagen-Dazs ice cream was selected for the control ice cream because of the high lipid content and factory quality control limits. Pasture-fed ice cream was procured from Picket Fence Creamery in Wellman, IA. The creamery indicated that the lipid content should be similar to Haagen-Dazs, though they could not guarantee with certainty. In-lab analysis indicated similar lipid content. Further analyses indicated that the Picket Fence ice cream may be as low as 14% lipid, while Haagen-Dazs remained close to 16% lipid.
Butter – Land-O-Lakes butter was used for the control butter and Pastureland butter was used for the treatment butter.
The 22 women that agreed to participate were started in a staggered fashion so that 12 participants started on Thursday and 10 on Saturday. This way of starting allowed for more personal interaction, which increases compliance, and also allowed for a more focused sample preparation and analysis. Of the 12 scheduled for Thursday, only 10 came to the first day of the study, of which all 10 completed the study. Of the 10 scheduled for Saturday, only 9 came to the first day of the study, of which one participant had to cease participation because of a medical concern unrelated to the study, leaving 8 participants that completed the study. In total, 18 women completed the entire study, of which 9 were in the control group and 9 were in the treatment group.
An oral glucose tolerance test (OGTT) was administered to fasting participants. Blood was collected immediately preceding consumption of a beverage containing 75g of glucose (Timepoint 0). The time at which the participant completed the beverage was recorded as time=0, after which blood was again drawn at time=30, 60, 90, 120, 150, and 180 minutes (Timepoints 1,2,3,4,5,6, respectively). Participants were given a light, treatment-neutral breakfast, and their treatments were subsequently assigned before lunch on the first day.
Participants were assigned to a treatment in the following manner: after the OGTT was completed, the serum cholesterol data from the participants were placed in numerical order. The participants were assigned to treatments in a pair-wise fashion such that the participant with the highest cholesterol was randomly assigned to one treatment and the participant with the next highest cholesterol was then assigned to the opposite treatment. Beginning treatment group statistics were then compared to confirm that no statistically significant differences existed between groups at the beginning of the study in either the Thursday group or Saturday group.
Meals were prepared for consumption in the Human Metabolic Unit dining room for breakfast, lunch, and dinner, Monday through Friday, and lunch on Sunday. The rest of the meals were packed out for the participants to bring home and consume. The balance between packed-out meals and eat-in meals was the result of attempting to maximize both participation and compliance, as more packed-out meals increases participation and more eat-in meals increases compliance. Food was weighed to the nearest tenth of a gram, with the exception of whole fruit and very-low calorie foods such as lettuce.
After 56 days on the study, participants again were given an OGTT, after which we celebrated their successful completion of participation and they were released from the study.
Serum samples were collected at day 1 and day 56 from each participant and were sent to LabCorp for clinical analysis. Plasma samples were collected in potassium EDTA tubes that were kept on ice until centrifugation at 4ºC for 15 minutes at 3000xg. Plasma was pipetted into several microcentrifuge-tube aliquots at Timepoint-0, two tubes for metabolite determination and two tubes expressly for glucose and insulin for OGTT analysis. Plasma was also aliquoted into glass tubes for glucagon analysis, as some theorize that glucagon will adhere to plastic tubes. Each subsequent timepoint had only two tubes for glucose and insulin, and one glass tube for glucagon. The sedimented red blood cell (RBC) fraction was saved and platelets were later separated by washing the RBC fraction with saline and centrifuging at 100xg.
For the analysis of lipoprotein fractions by the sequential flotation method of Havel et al., 3mL of plasma was pipetted directly into ultracentrifuge tubes.
DXA scans were conducted by a trained technician within the first few days of the study and within the last few days of the study.
Whole day meal samples were collected every day for each treatment in a manner that reflects the most a participant is likely to consume. That is, foods were scraped into the meal sample bags as though someone had eaten it, rather than rinsing the bowls into the bag. This method allowed for a more realistic approximation of consumption, and also allows for the calculation of total mandatory intake and total dry-matter intake.
-Methods of Analysis
Glucose concentrations were measured using a glucose Trinder assay. Insulin, adiponectin, and glucagon concentrations were measured using the appropriate commercially available radio-immuno assays (RIAs). Lipid content of food was measured using a modified Folch et al. lipid extraction procedure. Fatty acid profiles of the food were determined by methyl transesterification and subsequent analysis by gas chromatography. Dry matter content of the food was determined by freeze drying samples. Crude protein was approximated using a micro-Kjeldahl procedure using the freeze-dried samples.
An extra day’s diets for each treatment were collected daily including all provided foods and liquids. Daily total intake was determined gravimetrically. Samples were stored at 4ºC until homogenized by using a Waring industrial blender with a cooling coil (Waring Commercial 4L Laboratory Blender, 4L 2610T coil, New Hartford, CT) set to 4ºC by a cryogenic cooler (Caron Model 2050, Marietta, OH). Homogenate was frozen at -20ºC until analyzed for composition.
Daily dry matter intake of each subject was determined by lyophilizing 5 replicates of 5-g aliquots of daily intake samples. Protein was approximated by using a micro-Kjeldahl procedure to determine nitrogen content. Briefly, 60±3 mg of the dry matter sample were added to Kjeldahl tubes in duplicate; a Kjeltab (Thompson and Capper Ltd, UK) and sulfuric acid were added to each tube. Samples were digested for 4 hours, and titrated to determine total nitrogen. Protein was approximated by using a factor of 6.25 times the nitrogen content.
Daily total lipid consumption was determined by a modified Folch procedure (19) with 2:1 chloroform:methanol (v:v). Lipids were extracted in triplicate, and lipid content was determined gravimetrically.
Fatty acid methyl esters (FAMEs) were prepared by using sodium methoxide as a methanolic base and quantified by a gas chromatograph. Initial column temperature was set to 70ºC and held for 2 min; increased to 165ºC at 19ºC/min and held for 20 min; increased to 192ºC at 0.6ºC/min and immediately increased to 230ºC at 50ºC/min and held for 15 min for a total run time of 88 min. Samples were analyzed on a Varian 3350 equipped with a Varian 8200CX autosampler and a Supelco 2560 fused silica capillary column (100 mx0.25 mmx0.2 μm film thickness). FAMEs were identified by retention times compared with purified lipid standards (NuChek, Elysian, MN).
Trimethylsilyl derivatives of sterols were prepared from the non-saponifiable fraction of one of the triplicates of the lipid extract. Briefly, 5α-cholestane was added as an internal standard and the lipid extract was saponified in 11% (w/v) potassium hydroxide in a 55% (v/v) ethanol/deionized water solution. The nonsaponifiable fraction was isolated and derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide with 1% trimethylchlorosilane. Sterol derivatives were analyzed on a 30 m column with an initial column temperature of 150ºC that was increased to 200ºC for 20 min followed by an increase to 250ºC for 30 minutes. Select sterol derivatives were identified by retention times compared with free sterols derivatized as described.
All statistics were analyzed by using the general linear model procedure of SAS (version 9,1,; SAS Institute Inc., Cary, NC). Values were compared as either the magnitude or percentage change from day 0 to day 56.
Impacts and Contributions/Outcomes
The composition of the diets was reasonably close to the target composition (Table 2). The lower macronutrient energy percentages of protein and lipid is likely an artifact of assuming that all dry matter was protein, lipid, or digestible carbohydrate, which results in an inflated carbohydrate approximation, overestimated total calories, and underestimated protein and lipid. However, for comparison purposes, the values obtained indicate a fairly similar dietary composition, with no differences in calculated intake of total energy, carbohydrate, and protein (all p>0.1), with the exception of a small (~2%), but statistically significant, difference in lipid percentage (p<0.05). Cholesterol intake was similar between the two diets.
Fatty acid profiles of the diets were similar in saturated fatty acid (SFA), monounsaturated fatty acid (MUFA), and polyunsaturated fatty acids (PUFA) content (p>0.05), but with a significant difference in total CLA content and in specific isomers of CLA (p<0.0001) (Table 3). This resulted in an approximately 3.5-fold increase in CLA content in the CLA diet over the Control diet.
Changes in cholesterol concentrations in each lipoprotein fraction and fractions summed as the total cholesterol concentration recovered from the lipoprotein fractionation were analyzed (Figure 1A). All fractions were not different on day 0. Neither magnitude nor percentage changes in cholesterol concentrations from the beginning of the study to the end of the study between the two treatments were significantly different (p>0.05).
Triacylglycerol concentrations before and after the study were within normal ranges, with no significant differences at baseline. Changes from day 0 to day 56 were compared between the two diets (Figure 1B), with no significant differences (p>0.05) being observed.
Oral Glucose Tolerance
Plasma glucose, insulin, and glucagon concentrations were compared by using various measures of glucose intolerance including differences at each timepoint, as overall curves, as area-under-the-curve values (Figure 2A-C), and insulin resistance indices including homeostatic measurement of insulin resistance (data not shown). No markers of insulin resistance or glucose tolerance proved significant at p
Fasting adiponectin concentrations also were compared (Figure 2D). Fasting concentrations at day 0 were not significantly different. No significant differences in the percentage or magnitude change in adiponectin concentrations were observed between treatments.
The DXA scans of the left arm, right arm, trunk, left leg, right leg, and head were analyzed separately and as part of the subtotal (all parts except for the head) and total body compositions. No significant changes were noted in the separate body regions between the two treatments (data not shown). Of particular interest to health are subtotal and trunk compositions (Figure 3) to compare measures of whole body and abdominal, which also showed no significant differences between the two treatments (p>0.05).
Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were measured as markers of liver function (Figure 4). No differences were observed between the two treatments at baseline, and no differences were observed between the within treatment changes from day 0 to day 56.
The focus of the present investigation was the comparison of the effects of naturally occurring CLA in beef and dairy products of commercial versus pasture-fed origin. However, Table 3 indicates that there are a number of other dietary fatty acids with significant differences in concentration. In particular, the lipid percentages of identified odd-chain saturated and monounsaturated fatty acids were higher in the CLA diet than the Control diet. Little research has been done on these fatty acids with regard to human health. Furthermore, the ratio of identified n3:n6 fatty acids was significantly different (analysis not shown), and particularly arachidonic acid (C20:4-n6), a precursor of inflammation-related prostaglandins, was lower in the CLA diet and α-linolenic acid (C18:3-n3), an essential fatty acid related to decreased cardio-vascular disease, was higher in the CLA diet. These confounding factors can never be removed from a study of this nature because of the intended design of using a complete diet with no dietary supplementation.
AST and ALT, when elevated in concentration, are indicative of hepatic dysfunction. A number of studies have shown that high consumption of CLA alter risk factors associated with liver dysfunction (6-9). However, the present study agrees with the results presented by Iwata et al. that no differences occur in human models(10), indicating that the concentrations of CLA consumed resulted in no physiologically significant changes in liver function.
No alteration of body composition was noted between the two treatments. Abdominal adiposity is generally more closely associated with the morbidity of the metabolic syndrome, and previous CLA studies have shown alterations in abdominal adiposity with CLA supplementation. Once again, our results indicate that no such effect can be seen between the diets administered.
Glucose intolerance and insulin resistance, two terms that are often interchanged, did not show any difference between the two treatments, whether the comparison used the entire timecourse or just the baseline homeostatic comparison.
Every measurement taken on the participants fed the CLA diet showed no significant changes when compared with the participants fed the Control diet. The inconsistency of these data with some of the literature can be explained in a number of ways. First, the daily intake of CLA was lower than what some researchers have shown necessary to effect a change. Second, if there is a dramatic isomeric effect, c9,t11 CLA may not be the active form of CLA, and the t10,c12 CLA was in very low abundance in the diets. Third, there were a number of differences in other fatty acids in the diet besides CLA; if CLA does have negative effects on health, as some studies have shown, perhaps the improvement of other fatty acids with purported health effects counterbalanced these changes. Finally, this study was meant to compare two realistic diets with 9 human subjects fed each diet. Other human studies have shown effects on free living humans substituting products at home or taking encapsulated products, which can result in a number of confounding factors such as compliance and unmeasured changes in food consumption desire and patterns. Furthermore, these methods are able to provide CLA intakes well above that found in a natural diet. Perhaps the inconsistency of our data with those studies is that the physiological concentration of CLA necessary to effect change cannot, as of yet, be achieved naturally, and the changes may not be seen when compared with an appropriate control.
Diets high in beef and dairy products naturally enriched with CLA from pasturing the food animals do not result in improvements in markers of health in premenopausal women.
Resource Conservation & Development for NE Iowa In
101 E Greene, PO Box 916
Postville, IA 52175
Office Phone: 5638647112