|
|
||||||||
Original Research |
School of Biomedical Sciences, University of Tasmania, Tasmania (E.L.A., M.J.B.)
Department of Medicine, St. Vincents Hospital, Fitzroy, Victoria (J.D.B.), AUSTRALIA
Address reprint requests to: Madeleine J. Ball, M.D., School of Biomedical Sciences, University of Tasmania, PO Box 1214, Launceston, Tasmania 7250 AUSTRALIA. E-mail: Madeline.Ball{at}utas.edu.au
| ABSTRACT |
|---|
|
|
|---|
Design: A randomized crossover dietary intervention.
Setting: Free living individuals.
Subjects: Fourteen healthy males 35 to 55 years of age and 14 healthy postmenopausal women 50 to 60 years of age completed the dietary intervention. Two subjects did not complete the study, and their data were not included.
Interventions: A low fat, high carbohydrate diet (22% to 25% of energy from total fat, 7% to 8% of energy from monounsaturated fat and 55% to 60% of energy from carbohydrate) was compared to a monounsaturated enriched sunflower oil (MO) diet (40% to 42% of energy from fat, with 26% to 28% from monounsaturated fat and 40% to 45% of energy from carbohydrate) in an isocaloric substitution. Each dietary period was one month.
Results: Total cholesterol, LDL cholesterol, triglycerides and glucose were not significantly different between the two diets. HDL cholesterol, HDL3 cholesterol and insulin were significantly higher on the MO diet, mean 7%, 7% and 17% higher respectively. Copper-induced LDL oxidation lag phase was significantly longer (mean 18%) after the MO diet compared to the low fat, high carbohydrate diet. LDL particle size was not significantly different.
Conclusions: The significant increase in LDL oxidation lag phase and the significantly higher HDL cholesterol on the MO diet would be expected to be associated with a decrease in CHD risk.
Key words: CHD, MUFA, LDL oxidation, LDL size
| INTRODUCTION |
|---|
|
|
|---|
|
| METHODS |
|---|
|
|
|---|
Deakin University Ethics Committee approved the study protocol, and each subject gave written, informed consent.
Study Design
The study was designed to compare the effects of two diets using a randomized crossover design. Prior to the commencement of the study, subjects were given detailed verbal and written instructions on completing a weighed food record, using accurate scales or household measures when weighing was not possible. Subjects then recorded their habitual diet for four days, including one weekend day, so their usual energy intake could be estimated [6].
Subjects were randomly assigned to either the low fat, high carbohydrate diet or the MO diet first. Each diet period was for one month with a two-week washout period between, during which subjects returned to their habitual diet. Both diets were designed to be nutritionally adequate and isocaloric with 15% to 18% of energy from protein, 7% to 8% of energy from PUFA and 7% to 8% of energy from SFA, with energy content appropriate to maintain subjects body weight throughout the study period. The MO diet was designed to have 40% to 42% of energy from fat, with 26% to 28% from MUFA, and 40% to 45% energy from carbohydrate. The added MUFA in the MO diet was largely SunolaTM oil (Meadow Lea Foods Ltd, Mascot, Australia). The low fat, high carbohydrate diet was designed to have 22% to 25% of energy from total fat, with MUFA supplying only 7% to 8% of energy, and 55% to 60% of energy from carbohydrate, mainly from lemonade, yogurt and fruits. Meadowlea Sunflower oil was used in the low fat, high carbohydrate diet (composition given in Table 1). The two diets supplied different amounts of vitamin E, mainly due to the different amounts and types of oils. Vitamin E was not matched because foods that contain large amounts of vitamin E also contain fat and because supplemental vitamin E and food vitamin E do not necessarily exert the same effect on CHD risk [7]. Subjects weighed and recorded their dietary intake for seven days during the last week of each diet period, and the records were checked for completeness. These records were then analysed by Foodworks, (Version 1.05, 1997, Xyris Software, Queensland, Australia) using the NUTTAB 95 database. Vitamin E intake was calculated separately from the UK food composition tables [8].
Height and body weight were measured before starting the dietary intervention and weight was monitored, using electronic scales, at the end of each diet period. Venous blood was collected prior to commencing the study and at the end of each diet period, after an overnight fast and at approximately the same time. Serum and plasma aliquots were stored at -80°C for subsequent laboratory analysis.
Laboratory Analyses
Serum TC, TG, HDL cholesterol and plasma glucose were measured by enzymatic colorimetric tests using Boehringer Mannheim Gmbh calibrator, quality controls and reagents (Boehringer, Mannheim, Nunawading, Australia) on a Hitachi 704 autoanalyser (Hitachi, Tokyo Japan). LDL cholesterol was calculated using the Friedewald equation [9]. Using the calculated LDL, very low density lipoprotein (VLDL) cholesterol was calculated as TC- (LDL + HDL). HDL2 and HDL3, the major subfractions in the heterogenous HDL group of particles, were also measured using enzymatic methods and Boehringer kits. HDL3 was directly measured then subtracted from the total HDL cholesterol values to determine HDL2 concentration using a slight modification of the method of Widhalm and Pakosta [10].
Insulin was analysed by radioimmunoassay using the Linco Human Insulin Specific RIA Kit (Linco, Missouri, US), which included all standards, controls, human radioactive insulin (125I) and reagents.
LDL was isolated by ultracentrifugation, and copper-induced oxidative modification was measured spectroscopically using a method previously described [11]. The maximum diene formation was calculated from the molar extinction coefficient for conjugated dienes (29,500 L·mol-1·cm-1) as described by Abbey et al. [12].
LDL particle diameter was determined by gel gradient electrophoresis using commercially available 313% non-denaturing native gels (Gradipore, Sydney, Australia) using a regression plot derived from standards of known diameter (28 nm latex beads (Duke Scientific, Palo Alto, CA), and high molecular weight standards of thyroglobulin 17 nm and ferritin 12.2 nm (Pharmacia, Piscataway, NJ) [13].
The two oils, SunolaTM and Meadowlea Sunflower oil, used in the diets were analysed for their vitamin E content using the HPLC method described by Su et al. [14].
Statistical Analysis
Statistical analysis was performed using SPSS (Version 8.0.0, SPSS Inc., Chicago, IL). Paired t tests were used to investigate the differences between the two diets for normally distributed data (TC, LDL cholesterol and insulin), with log transformation for glucose values. Wilcoxon signed rank test was used to compare results at the end of the two diets for data that were not normally distributed (HDL cholesterol, HDL2, TG, oxidation lag phase, maximum diene formation rate, oxidation rate and estimated VLDL). General linear model (GLM) was used to investigate the carry over and order effects by treating both time and diet as fixed factors and identification number as a random factor [15]. Pearsons correlation was used to investigate the relationship between normally distributed data and Spearmans correlation for non-normally distributed data.
| RESULTS |
|---|
|
|
|---|
The self-recorded dietary intakes for the 28 subjects are given in Table 2. The percentage of energy derived from protein was significantly lower on the MO diet, but the actual amount in grams was not significantly different. SFA was slightly higher on the MO diet in the total group; however, there was no significant difference in SFA or PUFA between the two diets in the males. In females, there were minor differences in PUFA and SFA, with the PUFA being 0.8% of energy lower on the MO diet and SFA 1% of energy higher on the MO diet [16].
|
The lipids and lipoproteins measured are given in Table 3, and results for LDL oxidation, glucose and insulin are given in Table 4. The mean LDL lag phase was 18.5% longer after the MO diet compared to the low fat diet, while maximum conjugated dienes concentration and oxidation rate were 10.3% and 12.4% lower after the MO diet. The mean insulin concentration increased significantly by 17% on the MO diet. Mean HDL cholesterol and HDL3 were both 7% higher on the MO diet. There was no significant difference in TC, TG, LDL cholesterol, LDL size, glucose and glucose:insulin ratio between the two diets.
|
|
There was a significant correlation between the change in dietary MUFA and the change in TC (r = 0.76, p = 0.001), HDL cholesterol (r = 0.41, p = 0.032), HDL2 (r = 0.46, p = 0.014) after the two diets, but not with LDL lag phase, oxidation rate or LDL size. There was no significant correlation between the change in LDL particle size between the two diets and change in TC, LDL, HDL or TG concentrations. However, LDL particle size was inversely correlated with TG concentration at baseline (r = -0.597, p = 0.040), on the low fat diet (r = -0.433, p = 0.027) and on the MO diet (r = -0.470, p = 0.018). LDL particle size was also significantly correlated with HDL cholesterol concentration at baseline (r = 0.527, p = 0.006), on the low fat diet (r = 0.637, p < 0.001) and on the MO diet (0.612, p < 0.001).
VLDL cholesterol estimates for baseline, low fat, high carbohydrate diet and MO diet were 0.63 ± 0.36, 0.67 ± 0.29 and 0.57 ± 0.30 mmol/L respectively. VLDL cholesterol was significantly different between the low fat and MO diet (p = 0.042). The male and female subjects studied did not respond differently to the two diets for any parameter.
| DISCUSSION |
|---|
|
|
|---|
The significant difference in HDL cholesterol after the two diets may be due to an increase on the MO diet, a circumstance which is similar to that in several other studies that found a high MUFA diet increased HDL cholesterol compared to a low fat, high carbohydrate diet [17,18]. However, these findings have not been universal and other studies have found a high MUFA diet has no effect or negative effects on HDL cholesterol [1922]. Alternatively, HDL cholesterol may have been reduced on the low fat, high carbohydrate diet due to the decrease in total fat [1] or possibly due to the high glycemic index carbohydrates used to replace the MUFA [23]. Few studies have measured the effects on HDL subfractions, although the HDL2 subfraction is thought to have a stronger inverse relationship to CHD than HDL3 [24]. Both subfractions increased on the MO diet, although only the HDL3 subfraction significantly, so there is no evidence the results were due to a shift of cholesterol between subfractions.
In many studies a high carbohydrate diet induces hypertriglyceridemia; however, Parks and Hellerstein suggest in a recent review that there is no proof that hypertriglyceridemia induced by a low fat, high carbohydrate diet increases CHD risk [25]. Dietary factors that may influence TG and HDL levels on a low fat diet include the fiber, sucrose and starch content. In our study, the dietary fiber contents were similar, but both total sugars and starch were significantly higher on the low fat diet, and TG levels tended to be higher, although the difference was not statistically significant.
Many studies have compared the effects of high MUFA with low fat, high carbohydrate diets on TC and LDL [22,2628]. However a meta-analysis by Clarke et al. [29] indicates that replacement of carbohydrates with MUFA produced no significant effect on TC or LDL cholesterol, as seen in this study. Compared with baseline values, both diets were effective in lowering LDL cholesterol, suggesting that either diet could be associated with a decreased risk compared to the baseline diet. The low fat diet resulted in higher estimated VLDL cholesterol compared to the MO diet, and it has been suggested that high carbohydrate diets lead to a decrease in clearance of VLDL, rather than an increase in VLDL production [30].
Of the two studies that have compared the effect of a high MUFA diet with a low fat, high carbohydrate diet on LDL oxidation [4,5], OByrne et al. showed that the lag time before the formation of conjugated dienes was significantly longer both in the six women consuming the low fat diet and the five women on a low fat MUFA enriched diet compared to baseline, which was a high SFA diet. Diets were not well controlled, as the "control" low fat group showed a significant decrease in PUFA from baseline to the end of the diet, as did the experimental group. There was also a significant decrease in total fat and SFA in the experimental group. Berry et al. found that LDL isolated after a high MUFA diet was less prone to peroxidation, as measured by thiobarbituric acid reactive substances [4]. In our study the lag phase for LDL isolated after the MO diet was on average 12 minutes longer than after the low fat, high carbohydrate diet, also suggesting that the LDL were less susceptible to oxidative modification. Maximum conjugated dienes formed were significantly lower on the MO diet than on the low fat, high carbohydrate diet, perhaps suggesting a lower PUFA concentration within the LDL particle, as PUFA is the primary substrate of oxidation.
The endogenous antioxidant concentrations of the LDL particle can influence the LDL oxidation [31,32]. In the present study the antioxidants in the background diet were similar, but the increase in LDL lag phase may be due to a number of components in the SunolaTM, including the high MUFA content, the larger amounts of vitamin E supplied on the MO diet, their interactions or other unidentified components of the oil. Antioxidant and fatty acid analysis of the LDL particle would have provided insight into the changes in composition of the LDL, but would not have indicated the individual effects on LDL oxidation.
The subjects LDL particles in the present study were predominantly the larger, more buoyant subclass of LDL (diameter range 25.626.6 nm) [33]. Superko and Krauss [34] suggest that a threshold TG concentration exists (approximately 1.8064 mmol/L) where LDL particle size will change. However, only six subjects crossed this TG value, and these subjects were not statistically different from those that did not cross the threshold. There was an inverse relationship between the TG concentrations and LDL particle size, although there was no statistical difference in TG levels or LDL size between the two diets.
Insulin levels have been found to influence LDL oxidation [35,36]. Pelikanova et al. found that an induced state of hyperinsulinemia significantly decreased the lag time, suggesting enhanced LDL oxidizability, in healthy subjects [35]. Rifici et al. found that cell mediated oxidation of LDL was increased by incubation with supraphysiological insulin concentrations [36]. The mechanism or mechanisms of action of insulin on LDL oxidation are largely unknown; however, it is speculated to be through increased glucose utilisation by cells, resulting in increased O2- production which could initiate oxidation [37] or the conversion of cystine to thiols and the secondary generation of O2- by the reoxidation of thiols [38]. The slightly higher insulin levels on our MO diet were not associated with increased LDL oxidizability.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Some financial assistance and foods were provided from Meadowlea Foods and the Grains Research Development Corporation.
Received April 6, 2000. Accepted March 27, 2001.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. R. Archer, B. Lamarche, A. C. St-Pierre, J.-F. Mauger, O. Deriaz, N. Landry, L. Corneau, J.-P. Despres, J. Bergeron, P. Couture, et al. High Carbohydrate and High Monounsaturated Fatty Acid Diets Similarly Affect LDL Electrophoretic Characteristics in Men Who Are Losing Weight J. Nutr., October 1, 2003; 133(10): 3124 - 3129. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |