Alpha-linolenic acid (ALA) is the natural precursor of the cardioprotective long-chain n−3 fatty acids. Available data indicate a possible beneficial effect of ALA on cardiovascular disease (CVD), but the response of various CVD risk factors to increased ALA intake is not well characterized. The purpose of the present study was to examine the effect of increased ALA intake on blood pressure in man.
Design, setting, subjects and interventions:
We used a prospective, two-group, parallel-arm design to examine the effect of a 12-week dietary supplementation with flaxseed oil, rich in ALA (8 g/day), on blood pressure in middle-aged dyslipidaemic men (n=59). The diet of the control group was supplemented with safflower oil, containing the equivalent n−6 fatty acid (11 g/day linoleic acid (LA); n=28). Arterial blood pressure was measured at the beginning and at the end of the dietary intervention period.
Supplementation with ALA resulted in significantly lower systolic and diastolic blood pressure levels compared with LA (P=0.016 and P=0.011, respectively, from analysis of variance (ANOVA) for repeated measures).
We observed a hypotensive effect of ALA, which may constitute another mechanism accounting in part for the apparent cardioprotective effect of this n−3 fatty acid.
Greek Ministry of Development, General Secretariat for Research and Technology.
Cardiovascular disease (CVD) remains a leading cause of death in most industrialized countries. Consumption of long-chain polyunsaturated fatty acids (PUFA) of the n−3 series found in fish oils, predominantly eicosapentaenoic acid (EPA); 20:5n−3) and docosahexaenoic acid (DHA; 22:6n−3) has been associated with reduced CVD morbidity and mortality (He et al., 2004; Holub and Holub, 2004; Harper and Jacobson, 2005). However, an important issue for the scientific community relates to the amount and type of n−3 PUFA that should be recommended in the diet (Holub and Holub, 2004). An alternative source of n−3 PUFA is alpha-linolenic acid (ALA; 18:3n−3). ALA is the natural precursor of EPA and DHA and could, therefore, provide a readily accessible source of dietary n−3 PUFA that could be further desaturated and elongated in vivo, even though conversion of ALA into EPA and especially DHA in man is very low (Burdge, 2004; Burdge and Calder, 2005; Harper et al., 2006; Williams and Burdge, 2006). ALA competes with its n−6 counterpart linoleic acid (LA; 18:2n−6) for desaturation and chain elongation reactions.
Available studies to date have failed to provide convincing evidence that ALA can reproduce the effects of its longer-chain counterparts (Harris, 2005). Nonetheless, in a recent systematic review, a possible benefit of dietary ALA consumption in reducing sudden death and nonfatal myocardial infarction was deduced (Harper and Jacobson, 2005). Still, the underlying mechanisms for this favorable effect are poorly understood, as the influence of ALA on the various CVD risk markers is not well characterized: a recent meta-analysis concluded that increased ALA intake produces modest, if any, cardioprotective changes (Wendland et al., 2006). There is very little documentation on the influence of ALA on major CVD risk factors in man such as blood pressure; only a handful of clinical trials have been performed and produced equivocal findings (Wendland et al., 2006).
A number of observational studies have attempted to delineate the relationship between blood pressure and ALA. Results from these surveys have indicated that dietary ALA intake is not related (Bemelmans et al., 2000; Djousse et al., 2001; Djousse et al., 2003) or negatively associated (Salonen et al., 1987; Salonen et al., 1988; Djousse et al., 2005) with blood pressure levels. Others have used instead the content of ALA in plasma lipids or adipose tissue and, likewise, reported either an inverse association (Berry and Hirsch, 1986; Bemelmans et al., 2000) or no relationship (Riemersma et al., 1986; Simon et al., 1996; Grimsgaard et al., 1999). Another approach has been the supplementation of the diet with ALA and the prospective monitoring of blood pressure. Collectively, results from these clinical trials provide no evidence of a significant influence of ALA on blood pressure (Singer et al., 1986; Kestin et al., 1990; Bemelmans et al., 2002; Finnegan et al., 2003b; Wilkinson et al., 2005), though there are reports for both hypotensive (Singer et al., 1990) and hypertensive (Adam, 1989) effects. It is clear from the aforementioned studies that available data on the relationship between ALA and blood pressure are equivocal and lack consensus. The purpose of the present study, therefore, was to examine under carefully controlled conditions the effect of a 12-week dietary supplementation with ALA on blood pressure in dyslipidaemic patients.
Materials and methods
Subjects were recruited from the Department of Cardiology of Laiko Hospital, Athens, Greece, after screening by medical history, physical examination, electrocardiograph and laboratory analysis. Eighty-seven male volunteers aged 35 to 70 years, first diagnosed for dyslipidaemia, without evidence of CHD, were enroled. Entry criteria included a plasma total cholesterol concentration higher than 5.2 mmol/l (200 mg/dl) and/or a high-density lipoprotein (HDL) cholesterol concentration lower than 1.03 mmol/l (40 mg/dl). Individuals with evidence of infection or coexistent diabetes mellitus, renal, liver or inflammatory disease were excluded. Also, subjects taking lipid-lowering or antihypertensive drugs, habitually consuming more than 30 U of alcohol/week, smoking more than two packs of cigarettes per day or habitually exercising more than 6 h/week were excluded. None of the patients was taking dietary supplements of any kind (e.g., fish oil capsules) during the experimental period. All subjects were fully informed about the purpose and methods of the study and provided a signed consent. The study was approved by the Ethical Committee of Harokopio University, Athens, Greece.
This study was designed as a 12-week prospective, 2-arm, parallel group trial. Using a 2:1 process, subjects were randomly divided in two groups and assigned to dietary supplementation with either 15 ml of flaxseed oil per day containing ∼8 g of ALA (ALA group, n=59), or 15 ml of safflower oil per day containing ∼11 g of LA (LA group, n=28). These intakes resulted in a n−6/n−3 ratio of 1.3:1 in the ALA group and 13.2:1 in the LA group, whereas keeping total fat intake at ∼36% of total energy intake in both groups. Flaxseed and safflower oils were provided by Savant International (Savant Distribution Ltd, Leeds, UK); the composition of the oils is given in Table 1. The supplement was taken three times daily, one 5 ml teaspoon with each meal. In this study, in accordance with previous ones (Finnegan et al., 2003a, 2003b), the LA group was chosen to represent the control condition. Compliance was assessed by asking the subjects to return the bottles with unused oil for measuring the remaining volume; all participants consumed >98% of the assigned oil quantity.
Dietary assessment was performed at entry to confirm that all patients were following the average Greek diet. They were asked to maintain their dietary habits and usual lifestyle, and instructed to avoid intake of anti-inflammatory drugs, vitamins or any other dietary supplements throughout the intervention period. Subjects were supervised with respect to their dietary habits, alcohol consumption and physical activity behavior by phone calls once a week, and by visits to the hospital once a month. At the time of the monthly visits, subjects were weighed and were provided with the oil supplements. Dietary compliance was assessed by collecting one 3-day dietary record during each patient's monthly visit (total of three dietary records over the 12-week intervention period). The recorded days comprised of 2 weekdays and 1 weekend day. Dietary analysis was performed by using the Nutritionist V program (Version 2.1, First Data Bank Inc., San Bruno, CA, USA).
Blood pressure measurement
Arterial blood pressure was measured at the beginning and the end of the dietary supplementation period by using a standard manual sphygmomanometer, with the individual in a sitting position and after a resting period of at least 15 min. Blood pressure measurements were taken by a single physician, who was unaware of the oil supplement (ALA or LA) each patient was given. Three measurements were made at the right arm and averaged, with the arm relaxed and supported by a table at an angle of ∼45° from the trunk. The appropriate cuff size was used to accommodate different arm circumferences. Systolic blood pressure (SBP) was recorded at the first phase of the Korotkoff sounds, and diastolic blood pressure (DBP) was recorded at the fifth phase of the Korotkoff sounds. Mean arterial pressure (MAP) was calculated as MAP=((2 × DBP)+SBP)/3.
Normally distributed variables are reported as means±standard errors (s.e.), whereas data on blood pressure, which were not normally distributed even after logarithmic transformation, are expressed as the median and tertiles. Data were tested for normality with the Anderson–Darling test. Baseline data were compared between groups using the Student's unpaired t-test (for the normally distributed) or the Mann–Whitney U-test. Categorical variables were compared between groups at baseline with the χ2 test. Within-group comparisons, before and after the dietary supplementation, were made using the Wilcoxon signed ranks test. Spearman's correlation coefficient (ρ) was calculated to examine the association between changes in blood pressure and body mass index (BMI, in kg/m2) as the latter could be a potential confounder in the relationship between diet and blood pressure in man (Beilin, 1987). The effect of the dietary intervention on blood pressure measurements during the study course was assessed by analysis of variance (ANOVA) for repeated measurements, after adjusting for age and BMI. Statistical significance was set at P< 0.05, based on two-sided tests. Analysis was carried out with SPSS 13.0 for Windows SPSS Inc., Chicago, IL, USA).
The LA and ALA groups did not differ in any of the baseline anthropometric characteristics (Table 2). Moreover, BMI did not change during the dietary intervention period in either group (P>0.05). Also, there were no significant differences between groups in the quantity and quality of the background diet, as well as in the percentage of light smokers (all P>0.05); none of the participants reported modifying their dietary and smoking habits during the study. Energy and macronutrient intake, as well as consumption of individual n−3 PUFA, were not different between LA and ALA groups (Table 3).
The results for blood pressure measurements are shown in Table 4. The two groups did not differ at baseline in SBP, DBP, and MAP (all P>0.05). However, after the dietary supplementation, SBP, DBP and MAP were all significantly lower in the ALA group compared with the LA group (all P0.001). Within-group analysis showed that dietary supplementation with LA did not affect blood pressure levels (all P>0.05). On the other hand, dietary supplementation with ALA resulted in lower SBP, DBP and MAP than the respective baseline values (all P<0.001).
We further examined the relationship between baseline BMI and the blood pressure changes (post-intervention value minus baseline value, in mm Hg). In the LA group, δ SBP correlated significantly and negatively with baseline BMI (ρ=−0.444; P=0.018), as did δ MAP (ρ=−0.394; P=0.038), and a similar strong trend was noted for δ DBP (ρ=−0.356; P=0.063). On the other hand, we observed no significant associations between baseline BMI and δ SBP (ρ=0.049; P=0.714), delta DBP (ρ=0.059; P=0.659), or δ MAP (ρ=0.048; P=0.718) in the ALA group.
In the present study, we evaluated the effect of a 12-week dietary supplementation with flaxseed oil, rich in ALA, on resting blood pressure in dyslipidaemic patients. Our results indicate that increased ALA intake can bring about a significant decrease in SBP and DBP by approximately 5 mm Hg or 3–6% (medians). This blood pressure-lowering effect following the ALA-supplemented diet was significantly different than that following the control, LA-enriched diet.
Results from previous clinical trials do not support a beneficial effect of ALA. In the Mediterranean Alpha-linolenic Enriched Groningen Dietary Intervention study, Bemelmans et al. (2002) supplemented the diet of their subjects (middle-aged men and women at risk of CVD) with ALA (6.3 g/day) in the form of enriched margarine for a period of 2 years. At the end of the intervention period, no changes in SBP or DBP were observed (Bemelmans et al., 2002). In a placebo-controlled, parallel trial involving ∼30 moderately hyperlipidaemic subjects per arm, dietary supplementation with 4.5 or 9.5 g/day ALA for 6 months did not significantly affect SBP or DBP (Finnegan et al., 2003b). Similar results were reported by a handful of other studies of shorter duration. In such a trial with healthy men and women, hypertensive men, and hyperlipoproteinaemic men and women (n=10–15 per group), high amounts of ALA (∼38 g/day) for 2 week did not substantially affect SBP or DBP in either group (Singer et al., 1986). Wilkinson et al. (2005) carried out a study with some 20 healthy middle-aged males per arm, and supplemented their diet with ALA (∼15 g/day) for a period of 12 week; no changes in SBP or DBP were observed. In another investigation with 11 normotensive and mildly hypercholesterolaemic men per group, increased ALA intake (9.2 g/day) for 6 week did not influence SBP or DBP (Kestin et al., 1990). Slightly different results were obtained in a study with young adult male patients with mild essential hypertension, where 2 week of very high ALA (∼38 g/day) intake did not affect SBP or DBP at rest but, interestingly, resulted in ∼8% significantly lower SBP but not DBP during and after a psychophysiological stress test (Singer et al., 1990). On the contrary, in a small trial with normotensive females aged 23–32 years (n=6 per group), 6 week of increased ALA intake (0, 4, 8, 12 and 16% of total energy intake) produced almost stepwise increases in both SBP and DBP (Adam, 1989).
Against this background, we found that supplementation with ∼8 g/day ALA for 12 week significantly decreased both SBP and DBP in middle-aged dyslipidaemic men. The reasons why previous investigators failed to observe this blood pressure-lowering effect of ALA are unclear. Trials involving dyslidipaemic middle-aged adults (Singer et al., 1986; Kestin et al., 1990; Bemelmans et al., 2002; Finnegan et al., 2003b), administering similar and even higher ALA doses (Singer et al., 1986; Kestin et al., 1990; Finnegan et al., 2003b; Wilkinson et al., 2005), and of similar or even longer duration (Bemelmans et al., 2002; Finnegan et al., 2003b; Wilkinson et al., 2005) did not provide evidence of a hypotensive effect of ALA. Several confounding factors in studies of n−3 fatty acids and blood pressure have been identified and might have played a role, including the high variability of blood pressure in humans, the type of subjects and their selection, their habituation to blood pressure measurements, measurement errors and observer biases and multiple effects resulting from dietary changes (Knapp, 1989). Recently, for instance, it was shown that replacing saturated for monounsaturated fat in the diet lowers blood pressure, but this effect is negated at high total fat intakes (Rasmussen et al., 2006). All of these confounders could potentially obscure relatively small changes in blood pressure. In the present study, effort was made to control for most of these factors and minimize their influence. Subjects were carefully selected and monitored throughout the study with respect to their dietary, physical activity and smoking habits, blood pressure measurements were performed by specialized personnel according to standard procedures, and subjects were familiar with the clinic environment where the measurements were carried out.
The physiological mechanisms by which dietary ALA might lower blood pressure are not well understood. The most commonly held belief is that dietary fatty acids can modulate blood pressure through their effects on prostaglandin metabolism (Codde and Beilin, 1986; Knapp, 1989). In fact, there are many actions of prostaglandins relevant to blood pressure regulation, including direct effects on vascular reactivity, control of salt and water balance, regulation of renin release, effects on peripheral sympathetic tone, control of renal blood flow and effects on cardiac output and baroreceptors (Codde and Beilin, 1986; Knapp, 1989). Careful evaluation of both animal and human data, however, suggests that whereas n−3 fatty acids may indeed be able to lower blood pressure, they probably do not do so directly via altered production of prostaglandins (Knapp, 1989). Rather, the observed differences in prostaglandin biosynthetic capacities appear to be secondary to the blood pressure-modifying stimulus or to the change in blood pressure per se (Codde and Beilin, 1986). In any case, not only is our understanding of the involvement of the prostaglandin system in blood pressure regulation limited, but also, the effects of individual PUFA on prostaglandin metabolism are not well characterized: the formation of the various prostaglandins from fatty acid precursors varies considerably from one tissue to another, and also between species (Galli et al., 1980). To date, studies in man are very limited and have provided equivocal findings to allow for meaningful interpretation (Zollner et al., 1979; Adam et al., 1986). Future studies should be designed to address the underlying mechanisms through which dietary ALA might lower blood pressure.
There are some limitations of the present study that need to be addressed. First, we chose a parallel rather than a crossover design, despite the latter being inherently much stronger. This was done to avoid any carryover effect associated with crossing over, as the wash-out period cannot be readily specified in lipid intervention trials. Assuming an intermediate wash-out period of at least 2 months, the study would last almost 8 months, thus raising compliance issues. Second, the dose of ALA provided by the supplements (∼8 g/day) is difficult to be achieved by usual dietary intake (∼1.5 g/day) (Burdge, 2004). However, several products like cooking oil, margarine, salad dressing, and mayonnaise fortified with ALA can be produced by the industry, and inclusion of these foods in the diet has been shown to substantially increase dietary ALA intake to levels exceeding those used in the present study (∼9 g/day) (Mantzioris et al., 2000). Hence we believe our results could be applicable in practice.
In conclusion, dietary supplementation with 8 g/day ALA for 12 week lowers both SBP and DBP in dyslipidaemic men. The magnitude of the hypotensive effect (5 mmHg or 3–6%) is certainly clinically relevant, and is expected to considerably reduce the overall CVD risk in these patients. The blood pressure-lowering effect of ALA may be another mechanism accounting in part for the cardioprotective effect of this n−3 fatty acid.
We thank Dr E Karpodini, Dr E Theriou and the staff of the Department of Cardiology, Laiko Hospital, for assistance with subject management. This project was supported by the Greek Ministry of Development, General Secretariat for Research and Technology (grant no. 97EL-55).