Content uploaded by Graham A Colditz
Author content
All content in this area was uploaded by Graham A Colditz
Content may be subject to copyright.
ABSTRACT
Background: Experimental studies in laboratory animals and
humans suggest that a-linolenic acid (18:3n23) may reduce the
risk of arrhythmia.
Objective: The objective was to examine the association between
dietary intake of a-linolenic acid and risk of fatal ischemic heart
disease (IHD).
Design: This was a prospective cohort study. The intake of
a-linolenic acid was derived from a 116-item food-frequency
questionnaire completed in 1984 by 76283 women without pre-
viously diagnosed cancer or cardiovascular disease.
Results: During 10 y of follow-up, we documented 232 cases of
fatal IHD and 597 cases of nonfatal myocardial infarction. After
adjustment for age, standard coronary risk factors, and dietary
intake of linoleic acid and other nutrients, a higher intake of a-
linolenic acid was associated with a lower relative risk (RR) of
fatal IHD; the RRs from the lowest to highest quintiles were 1.0,
0.99, 0.90, 0.67, and 0.55 (95% CI: 0.32, 0.94; P for
trend = 0.01). For nonfatal myocardial infarction there was only
a modest, nonsignificant trend toward a reduced risk when
extreme quintiles were compared (RR: 0.85; 95% CI: 0.61, 1.19;
P for trend = 0.50). A higher intake of oil and vinegar salad dress-
ing, an important source of a-linolenic acid, was associated with
reduced risk of fatal IHD when women who consumed this food
≥5–6 times/wk were compared with those who rarely consumed
this food (RR: 0.46; 95% CI: 0.27, 0.76; P for trend = 0.001).
Conclusions: This study supports the hypothesis that a higher
intake of a-linolenic acid is protective against fatal IHD. Higher
consumption of foods such as oil-based salad dressing that pro-
vide polyunsaturated fats, including a-linolenic acid, may reduce
the risk of fatal IHD. Am J Clin Nutr 1999;69:890–7.
KEY WORDS Ischemic heart disease, diet, a-linolenic acid,
risk, Nurses’ Health Study, trans fatty acids, women
INTRODUCTION
Experimental studies have suggested an antiarrhythmic effect
of a-linolenic acid (18:3n23). In animal experiments in which
arrhythmias were induced by coronary occlusion and reperfu-
sion, significant reductions in the incidence of ventricular fibril-
lation and cardiac mortality were observed in rats fed with a-
linolenic acid–rich diets (1, 2). The antiarrhythmic effect of a-
linolenic acid may be attributable to an increased electrical
threshold for induction of ventricular fibrillation (1). In a sec-
ondary prevention trial of myocardial infarction (MI), patients
consuming the experimental diet with the higher a-linolenic
acid content experienced a significant reduction in cardiovascu-
lar death (3).
We examined the dietary intake of a-linolenic acid in relation
to the risk of fatal ischemic heart disease (IHD) among partici-
pants in the Nurses’ Health Study. Specifically, we conducted 2
parallel analyses: 1 to assess risk of fatal IHD among women
without prior diagnosed IHD at baseline and 1 limited to women
with a prior MI. Because a-linolenic acid is prone to oxidative
modification, the availability of antioxidants such as vitamin E
may be important for its biological effects. Thus, we hypothe-
sized a priori that the inverse association between a-linolenic
acid intake and fatal IHD would be stronger among vitamin E
supplement users. In addition, because trans fatty acids can
inhibit the desaturation of a-linolenic acid and may thus alter its
biological function (4), we examined whether intake of trans
fatty acids might modify the relation between a-linolenic acid
and risk of fatal IHD.
SUBJECTS AND METHODS
Study population
The Nurses’ Health Study cohort was established in 1976
when 121700 female, registered nurses aged 30–55 y and resid-
ing in 11 large US states completed a mailed questionnaire about
Dietary intake of a-linolenic acid and risk of fatal ischemic heart
disease among women1–3
Frank B Hu, Meir J Stampfer, JoAnn E Manson, Eric B Rimm, Alicja Wolk, Graham A Colditz, Charles H Hennekens,
and Walter C Willett
1From the Departments of Nutrition and Epidemiology, Harvard School of
Public Health; Channing Laboratory, and the Division of Preventive Medi-
cine, Department of Medicine, Brigham and Women’s Hospital and Harvard
Medical School, Boston.
2Supported by grants HL34594, CA40356, and DK 46200 and nutrition
training grant T32DK07703 from the National Institutes of Health.
3Address reprint requests to FB Hu, Department of Nutrition, Har-
vard School of Public Health, 665 Huntington Avenue, Boston, MA
02115. E-mail: Frank.Hu@channing.harvard.edu.
Received August 13, 1998.
Accepted for publication December 21, 1998.
See corresponding editorial on page 827.
Am J Clin Nutr 1999;69:890–7. Printed in USA. © 1999 American Society for Clinical Nutrition
890
by guest on February 7, 2011www.ajcn.orgDownloaded from
their medical history and lifestyle. Every 2 y, follow-up ques-
tionnaires were sent to the participants to update information on
potential risk factors and to identify newly diagnosed cases of
coronary and other diseases.
In 1980, a 61-item food-frequency questionnaire was included
to assess the intake of specific fats and other nutrients. In 1984,
the food-frequency questionnaire was revised to include ques-
tions about 116 individual food items, many of which had been
combined into food groupings in the original questionnaire. Sim-
ilar questionnaires were used to update dietary information in
1986 and 1990. Because the revised questionnaires contained
more food items that were critical for the assessment of a-
linolenic acid intake, we used data from the 1984 questionnaire
as baseline.
After ≥4 mailings, 81757 women returned the 1984 diet ques-
tionnaire. In the primary analyses, we excluded those who left ≥11
items blank, those with implausibly high or low scores for total
food and energy intakes [ie, <2761 kJ (660 kcal) or >14644 kJ/d
(3500 kcal/d)], and those with previously diagnosed cancer,
angina, MI, stroke, or other cardiovascular diseases at baseline to
avoid biased reporting of diet induced by the presence of these dis-
eases. The final 1984 baseline population for the analysis of fatal
IHD among women without prior IHD included 76283 women.
For the analysis of fatal IHD among women with a prior MI, we
included 1117 women who completed the 1984 questionnaire and
were diagnosed with a nonfatal MI at baseline or during follow-
up. The study was approved by the Human Research Committee
at Brigham and Women’s Hospital.
Dietary assessment
The reproducibility and validity of the food-frequency ques-
tionnaires were described in detail elsewhere (5). The correlation
between the calculated dietary intake of a fatty acid from the
food-frequency questionnaire and the proportion of the fatty acid
in adipose tissue was 0.34 (P < 0.001) for linolenic acid, 0.37
(P < 0.001) for linoleic acid, and 0.40 (P < 0.001) for trans fatty
acids (6). The correlation coefficients for linolenic acid intake
were 0.57 between the 1984 and 1986 questionnaires and 0.48
between the 1986 and 1990 questionnaires.
To calculate intakes of linolenic acid and other nutrients, a
commonly used unit or portion size for each food (eg, one egg or
one slice of bread) was specified and the participants were asked
how often on average during the previous year they had con-
sumed that amount. Nine responses were possible, ranging from
“never” to “≥6 times/d.” The intake of nutrients was computed
by multiplying the frequency of consumption of each unit of
food by the nutrient content of the specified portions. Composi-
tion values for linolenic acids and other nutrients were obtained
from the Harvard University Food Composition Database (com-
piled on 22 November 1993) derived from US Department of
Agriculture sources (7) and supplemented with manufacturer
information. Because the US Department of Agriculture data-
base has values only for total linolenic acid, we accounted for
g-linolenic acid in animal fats when computing intakes of
a-linolenic acid. a-Linolenic acid constitutes most of the total
linolenic acid in the diet, but only accounts for <60% of total
linolenic acid in beef fat (F Sacks, personal communication,
1998). We therefore subtracted 40% of the total linolenic acid
content in beef. [Linolenic acid is present in only small amounts
in beef and other animal fats (8), so analyses using total linolenic
acid or the adjusted value yielded near identical results.] In addi-
tion, we repeated the primary analyses using a-linolenic acid
from plant sources only.
Ascertainment of cases
The primary endpoint for this study was fatal IHD that
occurred after the return of the 1984 questionnaire but before
1 June 1994; nonfatal MI was a secondary endpoint. Deaths were
identified from the National Death Index, next of kin, or the
postal system. Using all sources combined, we estimated that
follow-up for the deaths was >98% complete (9). Fatal IHD was
defined as fatal MI if it was confirmed by hospital records or
autopsy, or if IHD was listed as the primary cause of death on the
death certificate and evidence of previous IHD was available.
The statement of the cause of death on the death certificates was
never relied on by itself as providing sufficient confirmation of
death due to IHD.
We requested permission to review medical records from
women who reported having a nonfatal MI on a follow-up ques-
tionnaire. The records were reviewed by study physicians, who
had no knowledge of the self-reported risk-factor status. Nonfa-
tal MI was confirmed if it met the criteria of the World Health
Organization of symptoms plus either diagnostic electrocardio-
graphic changes or elevated cardiac enzyme concentrations (10).
MIs that required hospital admission and for which confirmatory
information was obtained by interview or letter, but for which no
medical records were available, were designated as probable
(17% of total MI cases). We included all confirmed and probable
cases in the analyses because results were the same after proba-
ble cases were excluded.
Data analysis
For the analysis of fatal IHD among women without prior IHD
at baseline, person-years (the number of persons studied times the
number of years of follow-up) for each participant were calcu-
lated from the date of return of the 1984 questionnaire to the date
of the fatal IHD event, death, or 1 June 1994. Women with fatal
IHD or cancer noted on previous questionnaires were excluded
from subsequent follow-up; thus, the cohort at risk included only
those who remained free from cancer and were alive at the begin-
ning of each follow-up interval. For the analysis of fatal IHD
among women with a prior MI, person-years were calculated
from the date of return of the 1984 questionnaire to the date of the
fatal IHD event.
Women were grouped in quintiles of intake of a-linolenic
acid adjusted for total energy intake by using regression analy-
sis (11) based on the 1984 questionnaire. Incidence rates were
calculated by dividing the number of events by person-years of
follow-up in each quintile. The relative risk (RR) was computed
as the rate in a specific category of a-linolenic acid intake
divided by that in the lowest quintile, with adjustment for 5-y
age categories and smoking status [never, past, and current
(1–14, 15–24, and ≥25 cigarettes/d). A multivariate pooled
logistic model (12) was used to estimate the RRs and 95% CIs
of fatal IHD associated with various intakes of a-linolenic acid
and foods that were primary sources of a-linolenic acid in this
cohort, with simultaneous adjustment for other risk factors.
Tests of linear trend across increasing categories of a-linolenic
acid or food intakes were conducted by treating the categories
as a continuous variable and assigning the median intake for the
category as its value. Nondietary covariates, including age, cig-
arette smoking, body mass index (BMI), menopausal hormone
a-LINOLENIC ACID AND IHD 891
by guest on February 7, 2011www.ajcn.orgDownloaded from
use, alcohol use, multivitamin use, and vitamin E supplement
use, were updated biennially. Aspirin use was assessed in 1984
and 1988. Vigorous exercise was assessed in 1980. All dietary
variables were assessed in 1984, 1986, and 1990. Our primary
analyses used the baseline diet. We also conducted alternative
analyses using the updated dietary information (13). In the
updated analyses, we related incidence of fatal IHD to the
cumulative average a-linolenic acid intake from all available
questionnaires up to the start of each 2-y follow-up interval. In
the multivariate analyses of the association between a-linolenic
acid intake and risk of fatal IHD among prevalent MI patients,
we adjusted for the same covariates described above.
RESULTS
During 698750 person-years of follow-up, we documented
232 cases of fatal IHD and 597 nonfatal MIs. Among 1117
women with prevalent MI, 74 had fatal IHD during the follow-
up. The mean (±SD) daily intake of a-linolenic acid in 1984
was 1.10 ±0.45 g, or 0.57 ±0.16% of total energy intake.
Median daily energy-adjusted a-linolenic acid intake ranged
from 0.71 g in the lowest quintile to 1.36 g in the highest quin-
tile. Approximately 70% of a-linolenic acid was from vegetable
or plant sources (eg, salad dressings and vegetable oils) (Table 1).
On the basis of the 116 food items in the food-frequency ques-
tionnaire, the largest contributors (in absolute percentage) to the
overall intake of a-linolenic acid were mayonnaise or other
creamy salad dressing (16.7%); oil and vinegar (eg, Italian)
salad dressing (12.2%); margarine (6.8%); beef, pork, and lamb
as a main dish (6.5%); dairy products; and green-leafy vegeta-
bles such as iceberg lettuce.
Mean age and BMI, prevalence of current smoking, his-
tory of hypertension, parental MI before 60 y of age, current
postmenopausal hormone use, regular exercise, multivitamin
use, vitamin E supplement use, and aspirin use did not vary
appreciably according to intake of a-linolenic acid (Table 2).
a-Linolenic acid intake was correlated with intakes of linoleic
acid (r = 0.68), oleic acid (r = 0.56), and trans fatty acids
(r = 0.39), but not with n23 fatty acids from fish (r = 0.07).
892 HU ET AL
TABLE 1
Major contributors to the overall intake of a-linolenic acid based on the
1984 food-frequency questionnaire in the Nurses’ Health Study
Foods Value
%
Mayonnaise or other creamy salad dressing 16.7
Oil and vinegar salad dressing (eg, Italian) 12.2
Margarine 6.8
Beef, pork, or lamb as main dish 6.5
Milk 4.1
Cheese 3.9
Iceberg lettuce 2.0
Dark bread 1.6
Ice cream 1.6
Pie 1.4
Tomato sauce 1.4
Broccoli 1.3
TABLE 2
Relation of potential ischemic heart disease risk factors to energy-adjusted a-linolenic acids intake in 19841
Quintiles of a-linolenic acid intake
12345
(lowest, 0.71 g/d) (0.86 g/d) (0.98 g/d) (1.12 g/d) (highest, 1.36 g/d)
Current smokers (%) 24.4 23.7 23.5 23.5 25.6
History of hypertension (%) 22.1 21.6 20.8 22.0 21.5
History of diabetes (%) 3.0 2.8 3.0 3.0 3.4
History of hypercholesterolemia (%) 9.1 8.2 7.9 8.1 8.2
Parental MI before age 65 y (%) 14.5 14.5 14.5 14.9 14.9
Current postmenopausal hormone use (%) 19.5 18.4 19.6 19.4 19.6
Vigorous exercise ≥1/wk (%) 43.0 42.5 42.7 44.5 44.2
Multivitamin use (%) 39.7 37.4 37.0 35.7 35.2
Vitamin E supplement use (%) 18.5 16.3 16.4 16.9 17.4
Aspirin use ≥1/wk (%) 44.4 45.8 45.4 45.1 42.7
Age (y) 50.1 ±7.2250.0 ±7.2 50.2 ±7.2 50.4 ±7.2 50.8 ±7.1
Alcohol (g/d) 8.3 ±14.1 6.9 ±11.2 6.8 ±10.5 6.8 ±10.0 6.5 ±9.6
BMI (kg/m2) 24.8 ±4.7 25.0 ±4.7 25.0 ±4.7 25.1 ±4.8 25.2 ±4.8
Saturated fat (g/d) 20.0 ±4.3 21.8 ±4.1 22.5 ±4.3 22.9 ±4.5 23.4 ±5.0
Linoleic acid (g/d) 7.9 ±2.4 9.0 ±2.1 9.9 ±2.0 10.8 ±2.1 13.0 ±2.9
Oleic acid (g/d) 17.9 ±4.1 19.7 ±3.6 20.6 ±3.5 21.2 ±3.7 22.2 ±4.0
Eicosapentaenoic and docosahexaenoic acids (g/d) 0.17 ±0.16 0.17 ±0.15 0.17 ±0.14 0.18 ±0.15 0.20 ±0.17
trans Fatty acids (g/d) 2.7 ±0.9 3.2 ±0.9 3.4 ±1.0 3.6 ±1.0 3.9 ±1.1
Dietary vitamin C (mg/d) 143 ±75 137 ±62 134 ±58 133 ±57 134 ±61
Dietary vitamin E (mg/d) 5.8 ±4.4 6.0 ±3.0 6.3 ±2.7 6.7 ±2.7 7.6 ±2.8
Folate (mg/d) 394 ±246 381 ±225 377 ±220 377 ±220 390 ±241
Fiber (g/d) 16.3 ±5.4 16.4 ±4.8 16.3 ±4.5 16.4 ±4.6 16.6 ±5.1
Vegetables (servings/d)33.0 ±1.7 3.3 ±1.8 3.6 ±1.9 3.9 ±2.0 4.2 ±2.3
1MI, myocardial infarction.
2x
–±SD.
3A composite score of 28 vegetable foods.
by guest on February 7, 2011www.ajcn.orgDownloaded from
a-Linolenic acid intake was positively associated with intakes of
dietary vitamin E (r = 0.23) and vegetables (r = 0.22) and only
slightly, but significantly correlated with intake of vitamin C
(r = 20.04) and alcohol (r = 20.05).
Analyses of women with no prior IHD at baseline
After adjustment for age and smoking, energy-adjusted intake
of a-linolenic acid was inversely associated with risk of fatal
IHD; the RR for the highest compared with the lowest quintile
was 0.71 (95% CI: 0.47, 1.08; P for trend = 0.03) (Table 2). This
inverse association became stronger after adjustment for other
coronary risk factors (RRs from the lowest to highest quintiles:
1.0, 0.99, 0.90, 0.67, and 0.55; 95% CI: 0.32, 0.94; P for
trend = 0.01). Additional adjustment for vegetable intake did not
materially change the result when extreme quintiles were com-
pared (RR: 0.58; 95% CI: 0.33, 1.01; P for trend = 0.02).
a-Linolenic acid intake was only weakly associated with
reduced risk of nonfatal MI (Table 3). The multivariate RR for
extreme quintiles of intake was 0.85 (95% CI: 0.61, 1.19; P for
trend = 0.50). The association for total IHD (combining non-
fatal MI and fatal IHD) was intermediate between that for fatal
IHD and nonfatal MI when extreme quartiles were compared
(multivariate RR: 0.75; 95% CI: 0.56, 1.00; P for trend = 0.05).
In the multivariate model for fatal IHD, further adjustment for
intakes of fish n23 fatty acids, oleic acid, trans fatty acids,
cholesterol, folate, or fiber did not materially alter the associ-
ation. This inverse association was similar for aspirin nonusers
(RR for extreme quintiles of intake: 0.51; 95% CI: 0.24,1.09;
P for trend = 0.08) and users (RR for extreme quintiles of
intake: 0.60; 95% CI: 0.28,1.31; P for trend = 0.06). When
intake of a-linolenic acid was expressed as a percentage of
total energy intake rather than adjusted for total energy intake
by regression analysis, there was a similar inverse association
[RRs for fatal IHD from the lowest to the highest quintiles: 1.0
(reference group), 0.89 (95% CI: 0.59, 1.34), 0.90 (0.59, 1.38),
0.66 (0.41, 1.07), and 0.52 (0.30, 0.90); P for trend = 0.01].
The analyses limited to a-linolenic acid from plant sources
also yielded similar results [RRs for fatal IHD by quintiles of
a-linolenic acid intake: 1.0 (reference group), 1.17 (95% CI:
0.80, 1.74), 0.83 (0.52, 1.30), 0.55 (0.33, 0.94), and 0.64
(0.36, 1.13); P for trend = 0.03].
We conducted several alternative analyses of the relation of a-
linolenic acid intake to fatal IHD. The multivariate RRs for fatal
IHD when updated dietary information was used were 1.00, 0.96,
1.02, 0.71, and 0.57 (95% CI: 0.33, 0.98) from the lowest to high-
est quintiles of a-linolenic acid intake (P for trend = 0.02). When
we excluded women who had diabetes or hypercholesterolemia at
baseline, the RRs for fatal IHD by quintiles of a-linolenic acid
intake were 1.0, 0.93, 0.80, 0.58, and 0.55 (95% CI: 0.29, 1.07;
P for trend = 0.04). Exclusion of IHD cases that occurred in the
first 4 y of follow-up to avoid potential change in diet due to pre-
clinical conditions did not change the result.
We also examined the association between the ratio of
a-linolenic acid to linoleic acid and fatal IHD risk because
a-linolenic acid metabolism could be inhibited by high concen-
trations of linoleic acid (14). The median values for quintiles of
the ratio were 0.07, 0.09, 0.10, 0.11, and 0.14. The multivariate
RRs for quintiles of the ratio were 1.0, 1.02, 0.84, 0.87, and 0.84
(95% CI: 0.53, 1.33; P for trend = 0.40).
Stratified analyses
When the relation between a-linolenic acid and fatal IHD
was examined by use of vitamin E supplements (Figure 1), we
observed inverse associations among both users and nonusers
(P for interaction = 0.44), but a slightly stronger association
was seen among supplement users. When examined within cat-
egories of trans fatty acid intake, inverse associations between
a-linolenic acid and fatal IHD were observed among women
with higher and lower intakes of trans fatty acids (P for interac-
tion = 0.54), but the relation was somewhat stronger among
a-LINOLENIC ACID AND IHD 893
TABLE 3
Relative risks (RR) and 95% CIs of fatal ischemic heart disease and nonfatal myocardial infarction (MI) by quintiles of median energy-adjusted intakes of
a-linolenic acid
Quintiles of a-linolenic acid intake
12345Pfor
(lowest, 0.71 g/d) (0.86 g/d) (0.98 g/d) (1.12 g/d) (highest, 1.36 g/d) trend
Fatal CHD
No. of cases 49 52 51 41 39
Person-years 138468 139658 140606 139 711 140306
Crude mortality rate (per 100000 person-years) 35 37 36 29 28
RR adjusted for age and smoking 1.0 1.07 (0.73, 1.58) 1.02 (0.69, 1.51) 0.78 (0.52, 1.19) 0.71 (0.47, 1.08) 0.03
Multivariate RR11.0 0.99 (0.66, 1.48) 0.90 (0.59, 1.39) 0.67 (0.42, 1.09) 0.55 (0.32, 0.94) 0.01
Nonfatal MI
No. of cases 121 114 120 131 111
Person-years 138468 139658 140606 139 711 140306
Crude incidence rate (per 100000 person-years) 87 82 85 94 79
RR adjusted for age and smoking 1.0 0.95 (0.73, 1.22) 0.98 (0.76, 1.26) 1.05 (0.82, 1.35) 0.84 (0.64, 1.08) 0.29
Multivariate RR11.0 0.92 (0.71, 1.21) 0.94 (0.71, 1.25) 1.02 (0.76, 1.19) 0.85 (0.61, 1.19) 0.50
195% CIs in parentheses. Models include the following variables: age (5-y category), time period (7 periods), BMI (5 categories), cigarette smoking
[never, past, and current smoking (1–14, 15–24, and ≥25 cigarettes/d)], history of hypertension, history of diabetes, history of hypercholesterolemia,
menopausal status (premenopausal, postmenopausal without hormone replacement, postmenopausal with past hormone replacement, and postmenopausal
with current hormone replacement), parental history of MI before 65 y of age, multiple vitamin use, vitamin E supplement use, alcohol consumption (4 cat-
egories), aspirin use (nonuser, 1–6/wk, ≥7/wk, and dose unknown), vigorous exercise ≥1 time/wk, and dietary intakes (in quintiles) of saturated fat, linoleic
acid, vitamins C and E, and total energy.
by guest on February 7, 2011www.ajcn.orgDownloaded from
those with lower intakes of trans fatty acids (Figure 2). The
association between a-linolenic acid and fatal IHD risk did not
vary appreciably with intakes of linoleic acid or n23 fatty acids
from fish. In addition, the results were similar by smoking status
(current smokers versus nonsmokers) and alcohol drinking status
(current drinkers versus nondrinkers).
Analyses of women with a prior MI
We examined whether a higher a-linolenic acid intake could
reduce the risk of fatal IHD among women with a prior MI. After
adjustment for age, the RRs for fatal IHD across quintiles of
a-linolenic acid intake were 1.0 (reference group), 0.57 (95%
CI: 0.28,1.15), 0.77 (0.41,1.45), 0.48 (0.23, 0.99), and 0.51
(0.25, 1.06); the P for trend was 0.06. After adjustment for
nondietary and dietary confounding variables, the RRs were 1.0
(reference group), 0.61 (95% CI: 0.28, 1.30), 0.90 (0.41, 1.97),
0.61 (0.24, 1.52), and 0.35 (0.12, 1.01); the P for trend was 0.07.
Food analyses
To identify the foods that contributed most to differences in
a-linolenic acid intake among study participants, we used step-
wise regression with energy-adjusted a-linolenic acid as the
dependent variable and all foods as predictor variables. The
major determinants of a-linolenic acid were oil and vinegar
salad dressing and mayonnaise or other creamy salad dressings
(total r2= 0.50). We then examined the relation of these salad
dressings to risk of fatal IHD (Table 4). For both foods, we had
to combine adjacent categories of responses in the original ques-
tions to provide a sufficient number of women in each group.
Greater intake of oil and vinegar salad dressing was associated
with a reduced risk of fatal IHD (multivariate RR for the com-
parison of women who consumed the dressing ≥5–6 times/wk
and those who consumed it <1 time/mo: 0.46 (95% CI: 0.27,
0.76; P for trend = 0.001). A higher intake of mayonnaise or
other creamy salad dressing was associated with a nonsignificant
lower risk of fatal IHD [multivariate RR for the comparison of
women who consumed the dressing ≥5–6 times/wk and those
who consumed it <1 time/mo: 0.84 (95% CI: 0.50, 1.44; P for
trend = 0.44)]. Additional adjustment for vegetable intake did
not materially alter the results.
DISCUSSION
In this prospective study we found a significant inverse asso-
ciation between dietary intake of a-linolenic acid and risk of
fatal IHD. The risk was further lowered among women who also
took vitamin E supplements or who had a lower intake of trans
fatty acids. In addition, among women with prevalent MI, we
observed a trend toward lower risk of fatal IHD for those with a
higher intake of a-linolenic acid.
Inaccurate assessment of a-linolenic acid using a food-fre-
quency questionnaire is a potential concern. However, the calcu-
lated intakes of linolenic acid, linoleic acid, and trans fatty acids
were reasonably correlated with amounts of these fatty acids in
adipose tissue (6), suggesting the validity of our instrument in
measuring diet. In addition, we assessed the dietary intake of a-
linolenic acid and other nutrients 3 times during the study period,
so that analyses using updated dietary information accommodate
changes in dietary habits within individuals and in food composi-
tion. Confidence in the validity of our findings increased because
similar results were obtained when both baseline data and updated
dietary information were used. Women with a higher intake of a-
linolenic acid did not appreciably differ from those with a lower
intake for most of the known IHD risk factors, including age,
BMI, smoking, parental history of MI, postmenopausal hormone
use, and history of hypertension, diabetes, and hypercholes-
terolemia. This finding suggests that residual confounding by
these variables was unlikely to explain our findings. Women with
a higher intake of a-linolenic acid were more likely to have a
higher consumption of other dietary fats, especially linoleic acid.
However, adjustment for other fats actually somewhat strength-
ened the association. Because salad dressings are major sources of
894 HU ET AL
FIGURE 1. Multivariate relative risk (RR) of fatal ischemic heart disease by quintiles of a-linolenic acid intake and vitamin E supplement use in
women from the Nurses’ Health Study cohort. The model included the same variables as in Table 3. The reference group included women who were
in the lowest quintile of a-linolenic acid intake and nonusers of vitamin E supplements.
by guest on February 7, 2011www.ajcn.orgDownloaded from
a-linolenic acid, a-linolenic acid may be a proxy for vegetable
intake, but adjustment for vegetable intake did not alter the results.
In support of an inverse association between a-linolenic acid
intake and risk of fatal IHD, we found that specific foods that are
primary dietary sources of a-linolenic acid were associated with
reduced risk. In particular, we found a significantly lower risk of
fatal IHD among women who consumed oil and vinegar salad
dressing more frequently (≥5–6 times/wk) compared with those
who rarely consumed it. We also observed a nonsignificant inverse
association between intake of mayonnaise salad dressing and fatal
IHD. Both oil and vinegar and mayonnaise salad dressings are
commonly made from unhydrogenated soybean oil, which contains
<7% a-linolenic acid. However, the observed inverse associations
cannot be attributed to a-linolenic acid alone because these salad
dressings also contain linoleic acid and other potential beneficial
nutrients. Nevertheless, these data raise concern about widespread
use of fat-free salad dressings, which has eliminated an important
source a-linolenic acid and linoleic acid in the US diet.
a-Linolenic acid may contribute to a reduced risk of fatal IHD
through its antiarrhythmic effect. In cell culture studies, a-
linolenic acid was shown to slow the beating rate of isolated
neonatal rat cardiac myocytes (15). Also, significant reductions in
cardiac arrhythmia were observed in rats fed red meat supple-
mented with fish oil or canola oil (8% a-linolenic acid) when com-
pared with rats fed only red meat or red meat supplemented with
sheep fat (2). In another experiment, mortality due to ventricular
fibrillation was significantly lower in rats fed a diet containing
canola oil (0%) than in rats fed olive oil (30%) (1). In both exper-
a-LINOLENIC ACID AND IHD 895
TABLE 4
Relative risks (RR) and 95% CIs of fatal ischemic heart disease by intake of specific foods that are main sources of a-linolenic acid
Categories of intake
<1 time/mo 1–4 times/mo 2–4 times/wk ≥5–6 times/wk P for trend
Oil and vinegar salad dressing (15 mL)
No. of cases 84 92 38 18
Person-years 182719 265105 147431 103494
Crude mortality rate (per 100000 person-years) 46 35 26 17
RR adjusted for age and smoking 1.0 0.83 (0.62, 1.11) 0.60 (0.41, 0.87) 0.37 (0.22, 0.62) <0.001
Multivariate RR11.0 0.92 (0.68, 1.24) 0.68 (0.46, 1.01) 0.46 (0.27, 0.76) 0.001
Mayonnaise or other creamy salad dressing (15 mL)
No. of cases 28 108 66 30
Person-years 72935 320963 210238 94614
Crude mortality rate (per 100000 person-years) 38 34 31 32
RR adjusted for age and smoking 1.0 0.96 (0.63, 1.46) 0.88 (0.56, 1.36) 0.84 (0.50, 1.42) 0.37
Multivariate RR11.0 0.97 (0.64, 1.48) 0.91 (0.58, 1.43) 0.84 (0.50, 1.44) 0.44
195% CIs in parentheses. Models include the following variables: age (5-y category), time period (7 periods), BMI (5 categories), cigarette smoking
[never, past, and current smoking (1–14, 15–24, and ≥25 cigarettes/d)], history of hypertension, history of diabetes, history of hypercholesterolemia,
menopausal status (premenopausal, postmenopausal without hormone replacement, postmenopausal with past hormone replacement, and postmenopausal
with current hormone replacement), parental history of myocardial infarction before 65 y of age, multiple vitamin use, vitamin E supplement use, alcohol
consumption (4 categories), aspirin use (nonuser, 1–6/wk, ≥7/wk, and dose unknown), vigorous exercise ≥1 time/wk, and dietary intakes (in quintiles) of
saturated fat, linoleic acid, vitamins C and E, and total energy.
FIGURE 2. Multivariate relative risk (RR) of fatal ischemic heart disease by quintiles of a-linolenic acid and tertiles (T) of trans fatty acid intakes
in women from the Nurses’ Health Study cohort. The model included the same variables as in Table 3. The reference group included women who were
in the lowest quintile of a-linolenic acid intake and the lowest tertile of trans fatty acid intake.
by guest on February 7, 2011www.ajcn.orgDownloaded from
iments, arrhythmias were induced by coronary occlusion and
reperfusion. In a secondary prevention trial, cardiovascular deaths
were significantly lower among patients randomly assigned to a
Mediterranean diet enriched with a-linolenic acid than in those in
the control group, although other dietary changes occurred simul-
taneously in the experiment (3). In both animals and humans, a-
linolenic acid can be metabolized to long-chain polyunsaturated
n23 fatty acids, including eicosapentaenoic (20:5n23) and
docosahexaenoic (22:6n23) acids even though the efficiency of
this conversion is still a matter of debate (16). Previous studies
have suggested that higher intakes of long-chain n23 fatty acids
from fish may reduce the risk of fatal arrhythmias and sudden
death in humans (17, 18). Our observations that a-linolenic acid
were primarily associated with reduced risk of fatal IHD and that
higher a-linolenic acid intake reduced the risk of death among
prevalent MI patients are consistent with the antiarrhythmic effect
of a-linolenic acid. Our results are also consistent with those of
prospective studies in men in which the apparent protective effects
of a-linolenic acid were primarily seen for fatal IHD (19–21).
Besides its antiarrhythmic effect, a-linolenic acid may reduce
coronary risk through its antithrombotic effect (22). a-Linolenic
acid and its metabolite, eicosapentaenoic acid, can decrease gen-
eration of thromboxane A2, a proaggregatory vasoconstrictor,
through their inhibitory action on the conversion from linoleic
acid to arachidonic acid and the enzyme cyclooxygenase (15,
23–25). However, the similarity of the inverse association
between a-linolenic acid intake and risk of fatal IHD among
aspirin users and nonusers suggests that the effect of a-linolenic
acid is probably due to its antiarrhythmic rather than its
antithrombotic properties; otherwise, the effect of a-linolenic
acid would be masked by aspirin use.
It is thought that dietary a-linolenic acid may be essential
at 0.5% of energy intake (26). On the basis of food disap-
pearance data, Hunter (27) estimated the average dietary
intake of a-linolenic acid in the United States in 1985 to be <1.2
g/d, or 0.5% of the energy intake, which is similar to the mean
daily intakes estimated in the present study (1.1 g/d) and in the
Health Professional Follow-up Study (1.1 g/d) (20). These esti-
mates suggest that perhaps half the US population may not be
meeting the requirement for a-linolenic acid (16). In addition, the
optimal balance between dietary a-linolenic acid and linoleic
acid is not clear (1, 14). In this study, we found that the ratio was
less strongly related to risk of fatal IHD than was a-linolenic acid
alone because both fatty acids were inversely related to the risk.
We observed a somewhat stronger effect of a-linolenic acid
among vitamin E supplement users, suggesting that the avail-
ability of antioxidants such as vitamin E may be important for
the biologic effect of a-linolenic acid. In addition, because of its
susceptibility to oxidation, a-linolenic acid–rich oils are fre-
quently hydrogenated during processing, converting unsaturated
fatty acids to trans fatty acids. Substantial evidence suggests that
trans fatty acids have adverse effects on coronary disease (13,
20, 21, 28). In metabolic studies, trans fatty acids raise LDL and
lower HDL in humans (29). Moreover, trans isomers may inter-
fere with biological functions of essential fatty acids by compet-
ing with linoleic and a-linolenic acid for D6/5-desaturase (30). In
this study, we found that the reduction in fatal IHD risk associ-
ated with a-linolenic acid was somewhat attenuated among
women with a higher intake of trans fatty acids.
In conclusion, this study provides support for the hypothesis
that a higher intake of a-linolenic acid will reduce the risk of
fatal IHD. Our findings suggest that a reduction in consumption
of foods such as oil-based salad dressings that contain polyun-
saturated fats, including a-linolenic acid, may increase the risk
of fatal IHD.
We are indebted to the participants in the Nurses’ Health Study for their
continuing outstanding level of cooperation; to Al Wing, Mark Shneyder, Ste-
fanie Parker, Gary Chase, Karen Corsano, Lisa Dunn, Barbara Egan, Lori
Ward, and Jill Arnold for their unfailing help; and to Frank E Speizer, Princi-
pal Investigator of the Nurses’ Health Study, for his support.
REFERENCES
1. McLennan PL, Dallimore JA. Dietary canola oil modifies myocar-
dial fatty acids and inhibits cardiac arrhythmias in rats. J Nutr
1995;125:1003–9.
2. Siebert BD, McLennan PL, Woodhouse JA, Charnock JS. Cardiac
arrhythmia in rats in response to dietary n23 fatty acids from red
meat, fish oil and canola oil. Nutr Res 1993;13:1407–18.
3. de Lorgeril M, Renaud S, Mamelle N, et al. Mediterranean
a-linolenic acid-rich diet in secondary prevention of coronary heart
disease. Lancet 1994;343:1454–9.
4. Hill EG, Johnson SB, Lawson LD, Mahfouz MM, Holman RT. Per-
turbation of the metabolism of essential fatty acids by dietary hydro-
genated vegetable oil. Proc Natl Acad Sci U S A 1982;79:953–7.
5. Willett WC, Sampson L, Browne ML, et al. The use of a self-admin-
istered questionnaire to assess diet four years in the past. Am J Epi-
demiol 1988;127:188–99.
6. Garland M, Sacks FM, Colditz GA, et al. The relation between
dietary intake and adipose tissue composition of selected fatty acids
in US women. Am J Clin Nutr 1998;67:25–30.
7. US Department of Agriculture. Composition of foods: raw,
processed, and prepared, 1963–1988. Agriculture handbook no. 8.
Washington, DC: US Government Printing Office, 1989.
8. Nettleton JA. Omega-3 fatty acids: comparison of plant and seafood
sources in human nutrition. J Am Diet Assoc 1991;91:331–7.
9. Stampfer MJ, Willett WC, Speizer FE, et al. Test of the National
Death Index. Am J Epidemiol 1984;119:837–9.
10. Rose GA, Blackburn H. Cardiovascular survey methods. World
Health Organ Monogr Ser 1982;58.
11. Willett WC. Nutritional epidemiology. New York: Oxford Univer-
sity Press, 1990.
12. D’Agostino RB, Lee M-L, Belanger AJ, Cupples LA, Anderson K,
Kannel WB. Relation of pooled logistic regression to time dependent
Cox regression analysis: The Framingham Heart Study. Stat Med
1990;9:1501–15.
13. Hu FB, Stampfer MJ, Manson JE, et al. Dietary fat intake and risk of
coronary heart disease in women. N Engl J Med 1997;337:1491–9.
14. Chan JK, McDonald BE, Gerrard JM, Bruce VM, Weaver BJ, Holub BJ.
Effect of dietary a-linolenic acid and its ratio to linoleic acid on platelet
and plasma fatty acids and thrombogenesis. Lipids 1993;28:811–7.
15. Kang JX, Leaf A. Antiarrhythmic effects of polyunsaturated fatty
acids: recent studies. Circulation 1996;94:1774–80.
16. Cunnane SC. Metabolism and function of a-linolenic acid in
humans. In: Cunnane SC, Thompson LU, eds. Flaxseed in human
nutrition. Champaign, IL: AOAC Press, 1995:99–127.
17. Siscovick DS, Raghunathan TE, King I, et al. Dietary intake and
cell membrane levels of long-chain n23 polyunsaturated fatty acids
and the risk of primary cardiac arrest. JAMA 1995;274:1364–7.
18. Albert CM, Hennekens CH, O’Donnell CJ, et al. Fish consumption
and risk of sudden cardiac death. JAMA 1998;279:23–8.
19. Dolecek TA. Epidemiologic evidence of relationships between
dietary polyunsaturated fatty acids and mortality in the Multiple Risk
Factor Intervention Trial. Proc Soc Exp Biol Med 1992;200:177–82.
20. Ascherio A, Rimm EB, Giovannucci EL, Spiegelman D, Stampfer
MJ, Willett WC. Dietary fat and risk of coronary heart disease in men:
cohort follow up study in the United States. BMJ 1996;313:84–90.
896 HU ET AL
by guest on February 7, 2011www.ajcn.orgDownloaded from
a-LINOLENIC ACID AND IHD 897
21. Pietinen P, Ascherio A, Korhonen P, et al. Intake of fatty acids and
risk of coronary heart disease in a cohort of Finnish men: The Alpha-
Tocopherol, Beta-Carotene Cancer Prevention Study. Am J Epidemiol
1997;145:876–87.
22. Owren PA. Coronary thrombosis. Its mechanism and possible pre-
vention by linolenic acid. Ann Intern Med 1965;63:167–84.
23. Kinsella JE. Effects of polyunsaturated fatty acids on factors related
to cardiovascular disease. Am J Cardiol 1987;60:23G–32G.
24. Nair SSD, Leitch JW, Falconer J, Garg ML. Prevention of cardiac
arrhythmia by dietary (n23) polyunsaturated fatty acids and their
mechanism of action. J Nutr 1997;127:383–93.
25. Renaud S, Nordoy A. “Small is beautiful”: a-linolenic acid and eicos-
apentaenoic acid in man. Lancet 1983;1:1169 (letter).
26. Kinsella JE. Food lipids and fatty acids: importance in food quality,
nutrition, and health. Food Technol 1988;42:124–45.
27. Hunter JE. n23 Fatty acids from vegetable oils. Am J Clin Nutr
1990;51:809–14.
28. Willett WC, Stampfer MJ, Manson JE, et al. trans-Fatty acid intake
in relation to risk of coronary heart disease among women. Lancet
1993;341:581–5.
29. Mensink RPM, Katan MB. Effect of dietary trans fatty acids on
high-density and low-density lipoprotein cholesterol levels in
healthy subjects. N Engl J Med 1990;323:439–45.
30. Kinsella JE, Bruckner G, Mai J, Shimp J. Metabolism of trans fatty
acids with emphasis on the effects of trans, trans-octadecadienoate
on lipid composition, essential fatty acid, and prostaglandins: an
overview. Am J Clin Nutr 1981;34:2307–18.
by guest on February 7, 2011www.ajcn.orgDownloaded from
