Acute effects of monounsaturated fatty acids with and without omega-3 fatty acids on vascular reactivity in individuals with type 2 diabetes
- First Online:
- 911 Downloads
We examined the acute postprandial effects of meals containing unsaturated fatty acids on flow-mediated dilation (FMD) of the brachial artery and triacylglycerols in individuals with type 2 diabetes. We hypothesised that consumption of omega-3 fatty acids would enhance vascular function. Saturated fat reduces FMD for several hours, but there is inconsistent evidence about whether foods containing unsaturated fats impair FMD acutely. Little is known about the acute effects of omega-3 fatty acids on vascular reactivity.
We measured FMD before and 4 h after 3 test meals (50 g fat, 2,615 kJ) in 18 healthy adults with type 2 diabetes. The monounsaturated fatty acids (MUFA) meal contained 50 g fat from high oleic safflower and canola oils. Two additional meals were prepared by replacing 7% to 8% of MUFA with docosahexaenoic acid and eicosapentaenoic acid from sardine oil or α-linolenic acid from canola oil.
In the sample as a whole, FMD was increased 17% at 4 h vs. the fasting baseline. After the MUFA meal, subjects with the largest increases in triacylglycerols had the largest FMD decreases. The opposite pattern was observed after meals containing docosahexaenoic acid and eicosapentaenoic acid or α-linolenic acid. In subjects with high fasting triacylglycerols, meals containing 3 to 5 g of omega-3 fatty acids increased FMD by 50% to 80% and MUFA alone had no significant effects on FMD.
Endothelium-dependent vasodilation was not impaired 4 h after meals containing predominantly unsaturated fatty acids. The fatty acid composition of the meal and the metabolic status of the individual determine the vascular effects of a high-fat meal.
KeywordsDocosahexaenoic acid Eicosapentaenoic acid Endothelium-dependent vasodilation Flow-mediated dilation α-Linolenic acid Monounsaturated fatty acids Omega-3 fatty acids Type 2 diabetes mellitus
monounsaturated fatty acids
omega-3 fatty acids
saturated fatty acids
Dysfunction of the vascular endothelium is considered an important initial step in the development of atherosclerosis , and endothelial function is impaired in individuals with type 2 diabetes [2, 3]. Statins significantly improve endothelial function in healthy adults with high cholesterol [4, 5, 6], although it is questionable whether this strategy also works in patients with diabetes . In contrast, increased consumption of long-chain omega-3 (n-3) fatty acids is associated with reduced risk of cardiovascular disease in type 2 diabetes , and placebo-controlled studies suggest that improvements in endothelial function may be a mechanism for this effect .
Many previous studies have shown that a single meal containing high levels of saturated fat impairs endothelial function for 2 to 6 h (for review, see [10, 11]), and there is evidence that other nutrients contained in a meal can modify this response [12, 13]. Few studies have examined acute vascular effects of polyunsaturated and monounsaturated fatty acids (MUFA) [14, 15, 16]. The American Diabetes Association recommends that individuals with diabetes substitute MUFA for saturated fat in their diet , and oleic acid is the predominant unsaturated fatty acid in American diets. There is inconsistent evidence about whether diets high in oleic acid have a beneficial [5, 18, 19, 20], neutral [9, 15, 21] or negative [14, 16, 22] impact on markers of endothelial function, and few of the previously published studies were conducted in patients with diabetes.
Because olive oils can vary in their content of saturated and unsaturated fatty acids, we chose to measure the acute effects of blends of primarily unsaturated fatty acids. In this study, we directly compared the vascular and metabolic effects of fatty acid blends containing plant-derived and marine-derived n-3 fatty acids in adults with type 2 diabetes, and tested whether their inclusion in a meal altered the postprandial effects of oleic acid.
Subjects, materials and methods
Baseline characteristics (mean±SE) of participants
All subjects (n=18)
High triacylglycerols group (n=8)
Low triacylglycerols group (n=10)
Years since diabetes diagnosis
Fasting glucose (mmol/l)
Fasting insulin (μU/ml)
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
Total cholesterol (mmol/l)a
LDL cholesterol (mmol/l)
HDL cholesterol (mmol/l)
Resting artery diameter (mm)b
Flow-mediated dilation (Δ%)c
Resting arterial blood flow (ml/min)
Hyperaemic flow (Δ%)
Number (%) using oral hypoglycaemic medications
Potential participants were excluded if they had any of the following: fasting glucose >16.7 mmol/l; triacylglycerols >4.5 mmol/l; systolic blood pressure ≥160 mm Hg systolic; diastolic blood pressure ≥95 mm Hg; body mass index (kg/m2) >35; current smoker; history or current diagnosis of cardiovascular disease, diabetic retinopathy, neuropathy, or nephropathy; use of medications or supplements known to affect lipids or blood pressure. In addition to a physician’s examination and review of history, a 12-lead ECG and fasting blood chemistry panel were conducted. Tests for premenopausal women were scheduled during the early follicular phase (days 1–7) of the menstrual cycle. Written informed consent was obtained, and the protocol was approved by The Pennsylvania State University Institutional Review Board.
A randomised, double-blind, three-phase cross-over design was used. Treatment sessions were separated by at least 7 days. Participants were asked to abstain from alcohol for 48 h, and to discontinue all medications at 1800 hours the night before each test. Flow-mediated dilation (FMD) of the brachial artery was assessed in the right arm twice during each visit, once under fasting conditions and again 4 h after each meal. Blood samples were collected after a 12-h fast, and 2 h and 4 h after each meal. In previous studies, the largest changes in triacylglycerols and FMD have been observed at 4 h after the meal [23, 24, 25].
Composition of test meals
Protein (from skimmed milk)
Carbohydrate (from skimmed milk)
Assessment of vascular reactivity
The ultrasound protocol closely followed recommendations by Corretti et al. , and it has been described in detail previously . Brachial arterial diameter and flow velocity were measured using an Acuson 128XP duplex ultrasound imaging system (Siemens, New York, NY, USA) with a 10-MHz linear array transducer. Longitudinal, two-dimensional images of the brachial artery 5 to 10 cm above the elbow of the right arm were stored on SVHS tape during quiet rest (1 min), arterial occlusion via inflation of a cuff on the forearm (distal to the target artery) to 200 mm Hg (5 min), and reactive hyperaemia (2 min) by a sonographer with extensive training in vascular ultrasound. All images were of sufficient quality to be included in the analyses.
Frames for analysis were sampled at end diastole and diameters were measured continuously using automated edge-detection software (Brachial Analyzer; Medical Imaging Applications, Iowa City, IA, USA), with manual review of arterial boundaries by a trained technician, and confirmation by a second technician. FMD was measured as the percent change (Δ%) in arterial diameter from the average diameter under resting conditions to the peak diameter recorded during the deflation period. Blood flow was measured by spectral Doppler at rest and within 15 seconds of cuff release as flow (ml/min)=velocity time integral×cross-sectional area of the vessel×heart rate, and percent change in flow is reported here.
At screening, blood samples were assayed for lipids and lipoproteins, liver and kidney function, complete blood count, HbA1c, glucose, and insulin by American Medical Laboratories (Chantilly, VA, USA). Triacylglycerols were analysed using standard enzymatic procedures with commercially available kits (Abbott Laboratories, Diagnostic Division, Irving, TX, USA; Laboratory of Petar Alaupovic, Oklahoma Medical Research Foundation, Oklahoma City, OK, USA). Glucose was analysed via a portable enzymatic device (Precision, Abbott Laboratories, Medisense Products, Bedford, MA, USA). Serum insulin was measured using a solid-phase radioimmunoassay (Coat-A-Count diagnostic products, Los Angeles, CA, USA) based on insulin-specific antibodies and 125I-labelled insulin tracers . Insulin resistance and sensitivity scores  were calculated from fasting values of insulin and glucose as follows: HOMA-IR=glucose (mmol/l)×insulin (μU/ml)/22.5 and QUICKI=1/(log insulin×log glucose), where glucose is measured in milligrams per decilitre and insulin is measured in microunit per millilitre.
All variables were analysed using a mixed models approach and SAS PROC MIXED (SAS v. 8, Cary, NC, USA). Models included treatment, time and visit number as fixed effects, and subject as a random effect. There were no significant effects of treatment order. When significant effects of time were found, we examined the effect of treatment on change scores (postprandial level-fasted level). Tukey–Kramer adjusted p values were used to examine the source of significant effects, and all analyses of FMD and blood flow were adjusted for basal artery diameter. Text and tables report least squares means±SE, and probability values of ≤0.05 were considered statistically significant.
Interrelationships between the variables were estimated by partial Pearson correlation coefficients. Because fasting triacylglycerols concentration is an important predictor of postprandial metabolic and vascular response, we separately examined the effects of treatment in patients with high (≥1.69 mmol/l, n=8) vs. low (<1.69 mmol/l, n=10) fasting triacylglycerols, based on the average of four fasting values and utilising the National Cholesterol Education Program criterion to define high triacylglycerols . t Tests, analysis of covariance, and chi-square tests were used to compare the two groups on characteristics at study entry (see Table 1). Mixed models analyses were repeated as described above, with the triacylglycerols group and its interactions as additional fixed effects. In keeping with Herrington et al. , we calculated the coefficient of variation for fasting FMD for each participant as CV=SD×100/mean. Our CV was 29.7%, which is in the low range of published values .
Predictors of fasting flow-mediated dilation
As expected, lower fasting FMD was associated with increasing age (r=−0.72, p=0.001) and higher systolic blood pressure (r=−0.50, p=0.04). There was no relationship between fasting FMD and insulin, glucose, or markers of insulin sensitivity/resistance.
Vascular and metabolic effects of fatty acid blends
Effects of test meals on vascular and metabolic parameters
Resting artery diameter (mm)
Peak artery diameter (mm)
Resting blood flow (ml/min)a
Hyperaemic flow (Δ%)
Insulin (μU/ml) a
Predictors of postprandial change in FMD
Figure 1d–f shows the relationship between change in FMD and change in triacylglycerols at 4 h postprandial, separately for each of the three test meals. During the MUFA meal only, there was a significant inverse correlation between change in triacylglycerols and change in FMD (r=−0.50, p<0.05), such that subjects with the largest triacylglycerols increases after the MUFA meal exhibited the largest reductions in endothelium-dependent vasodilation. The opposite pattern was observed when test meals included 3 to 4 g of n-3 fatty acids such as ALA or EPA/DHA. During these test meals, larger increases in plasma triacylglycerols were associated with larger increases/improvements in FMD. The magnitude of this correlation was similar after consumption of EPA/DHA+MUFA (r=0.49, p=0.052) and ALA+MUFA (r=0.49, p=0.046). Postmeal changes in FMD were not correlated with any of the fasting metabolic or vascular parameters, nor with changes in insulin, glucose, arterial blood flow, blood pressure, or heart rate.
Effects of fasting triglyceride status on cardiac risk markers
Because it is well known that individuals with high fasting triacylglycerols also exhibit larger absolute changes in triacylglycerols following a high-fat meal, we compared postprandial responses in subjects with high vs. low fasting triacylglycerols. As shown in Table 1, individuals in the high triacylglycerols group were significantly older (p<0.05), had higher total cholesterol (p=0.05), and were taking more medications to control their diabetes than individuals in the low triacylglycerols group. There were no significant group differences in HbA1c, fasting insulin, glucose, HOMA-IR, or QUICKI under fasting conditions. There was a trend for the high triacylglycerols group to have smaller brachial artery diameter (p=0.08). When basal artery diameter was used as a covariate, fasting FMD was significantly lower in the high triacylglycerols group (p=0.03).
Group differences in postprandial responses to the meals
Effects of test meals on vascular and metabolic parameters in subjects having high or low fasting triacylglycerols
High fasting triacylglycerols
Low fasting triacylglycerols
Flow-mediated dilation (Δ%)a,b,c
Resting artery diameter (mm)
Peak artery diameter (mm)c
Resting blood flow (ml/min)a
Hyperaemic flow (Δ%)c
In contrast to previous studies showing impaired endothelial function after high-fat meals containing olive oil  or saturated fat [23, 32, 33, 34], FMD was not impaired 4 h after a high-fat meal containing predominantly unsaturated fatty acids. In fact, when the sample as a whole was considered, FMD significantly increased at 4 h. However, fasting triacylglycerols status, triacylglycerols response to the meals, and the fatty acid profile of the meals were important determinants of the direction and magnitude of postprandial responses. Subjects with elevated triacylglycerols concentrations showed significant FMD impairment under fasting conditions, slower clearance of glucose following a meal, and larger triacylglycerols increases after a meal containing 50 g of the MUFA fat blend when compared to the low triacylglycerols group. Importantly, however, when 7% to 8% of the MUFA fat blend was replaced with n-3 fatty acids from canola or sardine oil, the high triacylglycerols group exhibited significant increases in FMD and smaller increases in postprandial triacylglycerols concentrations (Fig. 2).
It is tempting to speculate that FMD improvements in the high triacylglycerols group in response to n-3 fatty acids were a direct result of the smaller increases in triacylglycerols that followed meals containing EPA/DHA and ALA. However, our data do not support a straightforward, causal link. During the MUFA meal there was a significant inverse correlation between change in triacylglycerols and change in FMD. In contrast, when test meals included either plant-derived or marine-derived polyunsaturated fatty acids, subjects with the largest increases in triacylglycerols had the largest improvements in FMD.
This apparent contradiction highlights an important issue in the study of triglyceride-rich lipoproteins: the composition of postprandial triacylglycerol particles generally reflects the fatty acid composition of the meal itself [35, 36]. Therefore, the increase in triacylglycerols after the MUFA meal may indicate higher circulating levels of oleic acid, which appears to be neutral with respect to FMD, while consumption of ALA or EPA/DHA would probably increase the concentration in serum. In regard to the latter, data from randomised, controlled, clinical trials have consistently shown that fish oil supplements improve endothelial function [9, 37, 38, 39], and two studies suggest that DHA may be the vasoactive component of fish oil [40, 41]. Recently, Leeson et al.  measured FMD and circulating fatty acid concentrations in 326 healthy young adults. They found a positive association between plasma levels of DHA and FMD scores and, consistent with findings reported here, this effect was evident in subjects with elevated triacylglycerols, but not in a control group with normal triacylglycerols. Khan et al.  found that fish oil increased the DHA content of erythrocyte phospholipids, and that subjects with the highest DHA concentrations after treatment had better endothelial function than subjects with low DHA concentrations. Our results should be confirmed after chronic treatment with n-3 fatty acids, and plasma concentrations of the bioactive fatty acids should be included in future studies.
In contrast to previous reports of FMD impairment following consumption of olive oil , we showed that FMD was not reduced following a meal containing 50 g of fat, including ∼30 g of oleic acid from high-oleic safflower oil. Our results confirm the findings of Williams et al. , who administered 64 g of olive oil or safflower oil, and found that neither source of unsaturated fatty acids had any acute effect on FMD, even after the oils were oxidised by heat. Given the mixed, and somewhat contradictory information about the effects of olive oil [14, 15], oleic acid [15, 16], and a Mediterranean diet [18, 21] on endothelial function, more work in this area is clearly needed.
Our finding that n-3 fatty acids increased FMD acutely also appears to be at odds with a study by Vogel et al. , who showed that a high-fat meal containing 50 g of fat from salmon had no impact on FMD. However, the sample tested by Vogel et al. included only healthy young adults with very low fasting triacylglycerols levels, and FMD improvements in the present study were only seen in subjects with high triacylglycerols. In addition, it is possible that other nutrients present in the fish altered the acute vascular effects of the n-3 fatty acids it contained.
We found no significant associations between postprandial changes in glucose, insulin, blood pressure, or heart rate, and change in FMD. However, because the peak insulin response occurred at 2 h, and FMD was not assessed until 4 h postmeal, we could not assess whether transient increases in insulin after a high-fat meal would affect vascular function. In addition, our enrolment criteria were designed to recruit a homogeneous sample of individuals with established diabetes but no clinically evident cardiovascular disease, and this may have excluded subjects with the largest postprandial changes in triacylglycerols, glucose and insulin. However, HbA1c levels in this sample ranged from 5.4% to 8.9% indicating that glucose control was not optimal despite the fact that the majority were taking one or more hypoglycaemic agents. These results cannot address whether patients with more advanced diabetes and/or cardiovascular disease would show a similar response to omega-3 fatty acids. Ros et al.  recently reported that FMD was significantly increased in healthy adults with hypercholesterolaemia after a mixed meal containing omega-3 fatty acids, suggesting that nondiabetic subjects may also benefit from including omega-3 fatty acids in their habitual diets.
Another study  proposed that the apparent reduction in FMD after a meal high in saturated fat is actually the result of peripheral vasodilation (measured as an increase in basal artery diameter), rather than a change in endothelial function in itself, and this pattern was confirmed in some [44, 45, 46] but not all [23, 33, 47] recent studies. In the present study, neither baseline artery diameter nor peak artery diameter were significantly changed after high-fat meals that are rich in unsaturated fatty acids. The meals had no significant impact on the magnitude of hyperaemic flow following cuff release, and adjustment for group differences in age and metformin status did not alter the pattern of vascular or triacylglycerols responses to the meals.
Assessment of FMD is not currently recommended as a clinical marker of cardiovascular disease risk, in part because small variations in technique have large effects on FMD estimates . We propose that triacylglycerols status be used to identify patients who may benefit from n-3 fatty acids, and our data suggest that marine and plant-derived n-3 fatty acids would be equally effective. Recent studies have found lasting improvements in FMD after longer term treatment with fish oil [9, 37, 38, 39], and with plant-derived ALA . For example, Ros et al.  showed that a fivefold increase in dietary ALA from walnuts increased FMD by 33%, even in the context of a Mediterranean-style background diet.
It is also important to consider the meaning of these effects in the context of recommended dietary intakes of n-3 fatty acids. Americans typically consume ∼1.6 g/d of n-3 fatty acids from plant and marine sources combined  and vegetable oils are the primary source. In this study, we provided 3.5 g of ALA and 4.8 g of EPA+DHA as part of a 50 g fat load. The National Academies  recommend that a nutritionally adequate intake of ALA is 0.8–1.1 g/day, and 10% of this amount may be supplied as EPA and DHA. The American Heart Association (AHA) Dietary Guidelines 2000 recommend consumption of at least two servings of fish (preferably oily) per week to decrease risk of cardiovascular disease , and this would typically provide approximately 350–450 mg of EPA and DHA per day (one serving of fatty fish, i.e., 85–113 g, provides approximately 1.2 to 1.6 g of EPA and DHA). The AHA Science Advisory “Fish consumption, fish oil, n-3 fatty acids, and cardiovascular disease”  recommends that patients with documented CHD consume approximately 1 g of EPA+DHA per day, preferably from oily fish. In addition, EPA+DHA supplements could be considered in consultation with the physician. The AHA acknowledges that for ALA, total intakes of approximately 1.5 to 3 g/day seem to be beneficial for cardiovascular disease risk reduction . This study could not address the effects of more moderate doses of omega-3 fatty acids. However, Ros et al. found improvements in FMD with plant-based omega-3 fatty acids within a meal that contained less fat (19–27 g) than the meals tested in the present study.
In summary, the present study showed that a meal containing 50 g of fat, primarily from unsaturated fatty acids, was not associated with impaired endothelial function. In patients with type 2 diabetes and high fasting triacylglycerols levels, meals containing 3 to 5 g of either plant or marine-derived n-3 fatty acids actually significantly improved postprandial lipaemia and endothelial function. Our data support the view that the vascular effects of a meal are dependent on both the fatty acid composition of the meal, and the metabolic status of the subject. From a clinical perspective, our results suggest that adjunctive treatment with n-3 fatty acids could enhance endothelial function even in patients who are already taking hypoglycaemic drugs. Future studies should assess whether the apparent vascular benefits of n-3 fatty acids are observed after long-term administration, and test whether changes in markers of oxidative stress, serum fatty acids, and l-arginine and nitric oxide metabolism are potential mechanisms for these effects.
Funding was provided by Ross Products Division, Abbott Laboratories, Columbus, Ohio, USA. Additional funding was provided by an Albert and Lorraine Kligman Graduate Fellowship awarded to K. D. Hecker, and from the General Clinical Research Center of The Pennsylvania State University (NIH #M01RR10732). We appreciate technical assistance provided by Matella Drum, Wake Forest University School of Medicine and Don Van Creveld of Acuson (Siemens Ultrasound Division), and Katrina LaRue Hurston at Penn State. Kari D. Hecker is now at Frito-Lay, Plano, TX, USA, and Stephanie L. Schoemer is in the Department of Human Nutrition, Michigan State University, Lansing, Michigan, USA. This study was funded by the Ross Product Division of Abbott Laboratories. The authors have stated that the results were not influenced by Abbott Laboratories.
- 2.Makimattila S, Yki-Jarvinen H (2002) Endothelial dysfunction in human diabetes. Curr Diabetes Rep 2:26–36Google Scholar
- 7.Van Venrooij FV, van de Ree MA, Bots ML, Stolk RP, Huisman MV, Banga JD (2002) Aggressive lipid lowering does not improve endothelial function in type 2 diabetes: the Diabetes Atorvastatin Lipid Intervention (DALI) Study: a randomized, double-blind, placebo-controlled trial. Diabetes Care 25:1211–1216PubMedGoogle Scholar
- 15.Williams MJA, Sutherland WHF, McCormick MP, Yeoman D, de Jong SA, Walker RJ (2001) Normal endothelial function after meals rich in olive or safflower oil previously used for deep frying. Nutr Metab Cardiovasc Dis 11:147–152Google Scholar
- 20.Ryan M, McInerney D, Owens D, Collins P, Johnson A, Tomkin GH (2000) Diabetes and the Mediterranean diet: a beneficial effect of oleic acid on insulin sensitivity, adipocyte glucose transport and endothelium-dependent vasoreactivity. Q J Med 93:85–91Google Scholar
- 21.Ambring A, Friberg P, Axelsen M et al (2003) Effects of Mediterranean inspired diet on blood lipids, vascular function and oxidative stress in healthy subjects. Clin Sci (Lond) 106:519–525Google Scholar
- 24.Ceriello A, Taboga C, Tonutti L et al (2002) Evidence for an independent and cumulative effect of postprandial hypertriglyceridemia and hyperglycemia on endothelial dysfunction and oxidative stress generation: effects of short- and long-term simvastatin treatment. Circulation 106:1211–1218CrossRefPubMedGoogle Scholar
- 26.Coretti MC, Anderson TJ, Benjamin EJ et al (2002) Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery: a report of the International Brachial Artery Reactivity Task Force. J Am Cardiol 39:257–265Google Scholar
- 30.National Cholesterol Education Program (2001) Third report of the expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). In: National Cholesterol Education Program. National Institutes of Health, National Heart, Lung, and Blood Institute, BethesdaGoogle Scholar
- 49.Institute of Medicine of the National Academies (2002) Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. In: Dietary Reference Intakes. The National Academies Press, Washington, DCGoogle Scholar