Intake of very-long-chain n-3 fatty acids related to social status

and lifestyle

LRK Johansson1, K Solvoll2, G-E Aa Bjùrneboe1 and CA Drevon1

1National Nutrition Council, Norway; and 2Institute for Nutrition Research, University of Oslo, Norway

Objectives: Little information is available about the intake of very-long-chain n-3 fatty acids in random samplesof populations. We examined if the intake of these fatty acids was associated with gender, social status andlifestyle in a similar way as other indicators for a healthy diet in a nationwide survey.Design and subjects: Data were obtained from self-administered quantitative food frequency questionnaires®lled in by a representative sample of Norwegian men and women, aged 16 ± 79 y. 3144 (63%) of the invitedsubjects responded with acceptable questionnaires.Results: Daily intake of very-long-chain n-3 fatty acids was on average 0.9 g=d and 0.4% of total energy wasderived from these fatty acids. Energy derived from very-long-chain n-3 fatty acids was slightly higher amongmen than women, and two-fold higher among subjects aged 60 ± 79 vs 16 ± 29 y. White collar workers had higherintake of very-long-chain n-3 fatty acids than blue collar workers. Men and women in the highest quartile ofintake of very-long-chain n-3 fatty acids had 2 ± 3 E% higher fat intake (mostly mono- and polyunsaturated fattyacids), as compared to individuals in the lowest quartile. They also had 3 ± 4 fold higher daily intake of retinoland vitamin D, as well as 20 ± 50% higher intake of fruits and vegetables, dietary ®bre and vitamin C.Conclusions: Intake of very-long-chain n-3 fatty acids was correlated to indicators for healthy dietary habits.However, contrary to many other indicators of a healthy diet, energy derived from very-long-chain n-3 fattyacids was not signi®cantly associated with female gender or non-smoking.Sponsorship: Ministry for Health and Social Affairs, Ministry for Agriculture and Norwegian Research Council.Descriptors: very-long-chain n-3 fatty acids; ®sh; cod liver oil; social status; lifestyle

Introduction

Very-long-chain n-3 fatty acids are biologically importantnutrients (Drevon, 1992). The main sources for these fattyacids are marine oils and ®sh. In some studies the intake ofmarine oils (Nestel, 1987) and ®sh (Kromhout et al, 1985;Daviglus et al, 1997) are associated with reduced risk forcoronary heart disease (CHD), although this has not beenobserved in other studies (Vollset et al, 1985; Morris et al,1995; Ascherio et al, 1995). Of four prospective studiesreporting intake of n-3 fatty acids from ®sh, one found aninverse association (Dolecek, 1992), another reported apositive association (Pietinen et al, 1997) and the othersreported no signi®cant association to CHD (Morris et al,1995; Ascherio et al, 1995). In two randomised clinicaltrials increased intake of ®sh (Burr et al, 1989) and alpha-linolenic acid (C 18:3, n-3) (de Lorgeril et al, 1994)reduced the mortality among patients who had recoveredfrom myocardial infarction, without signi®cantly affectingserum cholesterol concentration. In the DART trial adviceto eat at least two weekly portions of fatty ®sh wasassociated with a 29% reduction of total mortality in the®rst two years after myocardial infarction (Burr et al,

1989). In the other intervention study a dietary change toa Mediterranean alpha-linolenic acid-rich diet was asso-ciated with a 70% reduction of total mortality, providing0.6 g=d of alpha-linolenic acid in the control group and2 g=d in the intervention group (de Lorgeril et al, 1994). Asconversion of alpha-linolenic acid to eicosapentaenoic acid(EPA) occurs in humans, increased intake of alpha-linole-nic acid may have effects similar to those of very long-chain n-3 fatty acids from ®sh (McKeigue, 1994). Thus,these observations suggest that bioavailable very-long-chain n-3 fatty acids may be important for prevention ofCHD.

The intake of very-long-chain n-3 fatty acids isdescribed in clinical trials (Bùnaa et al, 1990; Sùyland etal, 1993; Sùyland et al, 1994; Brude et al, 1997), selectedgroups (Bang et al, 1980; Tjùnneland et al, 1993; Nydahl etal, 1996; Hjartaker et al, 1997) and prospective studies(Morris et al, 1995; Ascherio et al, 1995; Dolecek, 1992;Pietinen et al, 1997), but little information is availableabout the intake of very-long-chain n-3 fatty acids inrandom samples of populations (Steingrimsdottir et al,1995). We therefore assessed the intake and sources ofvery-long-chain n-3 fatty acids and related the intake tosocio-demographic and life-style variables in a nationaldietary survey. Healthy lifestyle is associated with highsocial status (Blaxter, 1990; PraÈttaÈlaÈ et al, 1994) andwomen usually have more healthy diets than men (Klee-mola et al, 1994; Johansson et al, 1997a). In the presentstudy we tested the following hypotheses:

Correspondence: Lars RK Johansson, National Nutrition Council, Box8139 Dep., N-0033 Oslo, Norway.Received February 26 1998; revised June 11 1998; accepted June13 1998

European Journal of Clinical Nutrition (1998) 52, 716±721ß 1998 Stockton Press. All rights reserved 0954±3007/98 $12.00

http://www.stockton-press.co.uk/ejcn

1. Women have higher intake of very-long-chain n-3 fattyacids than men.

2. High social status is associated to higher intake of very-long-chain n-3 fatty acids than low social status.

3. Subjects with healthy lifestyle (for example regularphysical exercise and non-smoking) have higher intakeof very-long-chain n-3 fatty acids than those withoutthese habits.

4. High intake of very-long-chain n-3 fatty acids is asso-ciated with indicators of a healthy diet.

Methods

Subjects

In a national dietary survey (NORKOST), a self-adminis-tered quantitative food frequency questionnaire was ®lledin by a representative random sample of Norwegians aged16 ± 79 y; 3144 subjects (63%) completed the questionnaire.The distribution of subjects in different groups of socio-economic status, location of residence and level of educa-tion was similar among our participants as compared to thegeneral population (Statistics Norway, 1995). Moreover,there were only small differences between responders andthe total random sample regarding age, sex, geographicaldistribution and educational level (Johansson et al, 1997a).

QuestionnaireThe questionnaire was designed to cover the whole diet. Itincluded about 180 food items and the frequency of con-sumption was given per day, per week or per monthdepending on the food item. The portion sizes werequanti®ed as slices, glasses, cups, pieces, decilitres andspoons and converted to weights on the basis of standardportions (Blaker & Aarsland, 1989). Questions aboutweight, height, physical activity, smoking habits, mealfrequency and personal attitudes related to diet and bodyweight, were also included in the form. Statistics Norwayprovided information about the subjects' level of educationand several other demographic and geographical variablesfrom their registers. Occupation was classi®ed as bluecollar workers (unskilled and skilled workers, and lowerlevel salaried employees) and white collar workers (meanand higher level salaried employees) (Central Bureau ofStatistics, 1984). Smoking habits were classi®ed as non-smokers or smokers of cigarettes or pipes. Frequency ofexercise was evaluated by the question: How often do youhave physical exercise for at least 20 min (walking, jog-ging, bicycling or swimming)? (Never, < 1, 1, 2 ± 3, 4 ± 6times=week or daily). Attention to healthy diet was deter-mined by the question: What attention do you pay tokeeping a healthy diet? (very low, low, medium, high orvery high). A more detailed description of the participants,the questionnaire, the calculation of nutrients and reporteddietary habits, is given in recent publications (Johansson etal, 1997a,b).

The questionnaire included questions about the fre-quency for use and portion size of cod liver oil, cod liveroil capsules and ®sh oil capsules. Use of cod liver oil hastraditionally been recommended during the dark period ofthe year in Norway. Therefore, questions about use of codliver oil was differentiated between the whole year and thewinter season only. Reported frequency for use during thewinter was divided by a factor of two.

Daily intake of energy and nutrients was computed usingthe 1995 version of the Norwegian food composition tableand software systems developed at Institute for NutritionResearch. The calculated intake of energy and nutrients inthe present analysis differs slightly from earlier publishedresults from this survey, since the conversion factor forenergy derived from fat was changed from 38 ± 37 kJ=g, anddietary ®bre was not included in the carbohydrates in the1995 as compared to the 1991 version of the food composi-tion table.

Total intake of very-long-chain n-3 fatty acids wascomputed as the sum of eicosapentaenoic acid (EPA)(20:5, n-3), docosapentaenoic acid (DPA) (22:5, n-3) anddocosahexaenoic acid (DHA) (22:6, n-3). Intake of very-long-chain n-3 fatty acids reported by the questionnaire haspreviously been evaluated against the concentration ofplasma phospholipids among 579 men and women (FrostAndersen et al, 1996). Correlation coef®cients betweenplasma content and dietary intake of fatty acids were 0.55and 0.56 for EPA and DHA, respectively, when intake andplasma concentration were expressed as percent of totalfatty acids. Other nutrients reported by the questionnairewere evaluated against 14 d dietary records among 38elderly women (Nes et al, 1992) and against 48 h recallamong 123 dermatologic outpatients (Solvoll et al, 1993).These studies showed an acceptable accuracy of the ques-tionnaire in assessing individual intakes of a wide range ofnutrients, including very-long-chain n-3 fatty acids.

StatisticsIntake of very-long-chain n-3 fatty acids was skewed andtherefore non-parametric statistical methods were chosenfor analysis. Differences between two or more groups weretested with Mann ± Whitney U test and Kruskal ± Wallis Htest, respectively. Spearman rank correlations were used totest the association between the intake of very-long-chainn-3 fatty acids, ®sh and cod liver oil. Chi square test wasused to test differences in proportion between subgroups ofparticipants. Data were analysed using the program SPSS(SPSS, 1996).

Results

Intake and sources for very-long-chain n-3 fatty acidsFish and cod liver oil contributed with 56% and 33%,respectively, of the total intake of very-long-chain n-3 fattyacids (Table 1). The distribution of the intake of very-long-chain n-3 fatty acids was skewed and the range for intakewas large. For all subjects mean and median daily intake ofvery-long-chain n-3 fatty acids was 0.89 and 0.59 g=d,

Table 1 Dietary sources of very-long-chain fatty acids, % of total intake,and absolute daily intake; means (s.d.)

EPAa DPAb DHAc Sum of EPA,20:5 22:5 22:6 DPA, DHA

Cod liver oil (%) 39 14 32 33Fish oil (%) 3 Ð 2 2Fish (%) 55 43 57 56Meat (%) 3 43 4 6Other foods (%) Ð Ð 4 2Total (%) 100 100 99 99Intake (g=d) 0.33 (0.40) 0.07 (0.05) 0.49 (0.51) 0.89 (0.95)

aEPA� eicosapentaenoic acid; bDPA� docosapentaenoic acid;cDHA� docosahexaenoic acid.

Intake of very-long-chain n-3 fatty acidsLRK Johansson et al

717

respectively, and mean and median dietary content was0.36 and 0.24 E%, respectively. Mean intake of ®sh was67 g=d. Less than 2% of the population reported no intakeof ®sh and 6% had an intake below 10 g=d. Cod liver oilwas never used by 65% of the subjects, whereas 19% usedit during the whole year, and only during the winter seasonby an additionally 16%. Mean intake of cod liver oil was1.3 g=d. Ten percent of the participants had a cod liver oilintake of 1 ± 4 g=d and an additional 10% had at least 5 g=d.Fish oil supplements were never used by 96% of thesubjects.

Spearman correlation coef®cients between intake ofvery-long-chain n-3 fatty acids vs ®sh intake and vsintake of cod liver oil were 0.61 and 0.57, respectively,and 0.17 between intake of ®sh vs cod liver oil (P< 0.001).Information about intake of cod liver oil and ®sh was mostimportant to classify subjects into quartiles of intake ofvery-long-chain n-3 fatty acids. Half of the subjects report-ing use of cod liver oil had an intake of very-long-chain n-3fatty acids within the range of the highest quartile and 8%within the lowest. Among subjects reporting use of codliver oil and belonging to the highest quartile of ®sh intake,72% had an intake of very-long-chain n-3 fatty acids withinthe range of the highest and < 1% had an intake within thelowest quartile of intake.

Intake of very-long-chain n-3 fatty acids related to genderand ageDaily intake of very-long-chain n-3 fatty acids, as well asother polyunsaturated fatty acids, was higher among mencompared to women (Table 2). The percent of total dietaryenergy (E%) from very-long-chain n-3 fatty acids wasslightly, but signi®cantly higher among men than women.Men also had higher daily intake of ®sh and cod liver oilthan women. However, when intake was computed per

10 MJ, men no longer had higher intake of ®sh and codliver oil than women. The fraction of male users of codliver oil was similar to female users (37% vs 34%). A largerfraction of men reported having the largest portion size ofcod liver oil (one table spoon=time) as compared to women(76% vs 58%). The mean frequency for use of cod liver oilwas 3.7 times=week among users of cod liver oil.

The percent of dietary energy derived from very-long-chain n-3 fatty acids, as well as the intake of ®sh and codliver oil, was higher in the older compared to the youngerage groups (Figures 1, 2, 3). Subjects aged 60 ± 79 y hadtwo-fold higher intake of very-long-chain n-3 fatty acids,®sh and cod liver oil than subjects aged 16 ± 29 y.

Characteristics of subjects with low vs high intake of very-long-chain n-3 fatty acidsMen and women in the highest quartile of intake of very-long-chain n-3 fatty acids had 10 times higher intake of

Table 2 Daily intake of energy, fatty acids, ®sh and cod liver oil, andfrequency for use of cod liver oil (times=week) among men (1517) andwomen (1627); means (s.d.)

Men Women P-valuea

Energy (MJ=d) 11.3 (4.1) 8.0 (2.8) < 0.001

Fat (g=d) 97.4 (44.4) 67.0 (29.1) < 0.001Fatty acids (g=d)Linoleic acid 18:2 n-6 13.5 (8.5) 8.8 (5.1) < 0.001Aracidonic acid 20:4 n-6 0.17 (0.08) 0.12 (0.05) < 0.001Linolenic acid 18:3 n-3 1.8 (1.2) 1.2 (0.7) < 0.001EPA 20:5 n-3 0.41 (0.46) 0.27 (0.31) < 0.001DPA 22:5 n-3 0.08 (0.06) 0.06 (0.04) < 0.001DHA 22:6 n-3 0.59 (0.59) 0.40 (0.40) < 0.001EPA, DPA, DHA, total 1.08 (1.1) 0.72 (0.75) < 0.001Ratio n-3=n-6 0.23 (0.13) 0.24 (0.14) 0.267

Percent of energy (%)Fat 31.5 (5.8) 30.6 (5.9) < 0.001SFA 12.3 (2.8) 12.2 (2.7) 0.597MUFA 11.1 (2.2) 10.7 (2.2) < 0.001PUFA 5.7 (1.9) 5.3 (1.8) < 0.001EPA, DPA, DHA, total 0.37 (0.39) 0.35 (0.37) 0.046

Fish (g=d) 78.(62) 57.(40) < 0.001Fish (g=10 MJ) 73.(55) 75.(52) 0.047Cod liver oil (g=d) 1.4 (3.0) 1.1 (2.6) 0.028Cod liver oil (g=10 MJ) 1.4 (3.1) 1.5 (3.7) 0.152Cod liver oil (times=week) 1.3 (2.2) 1.3 (2.2) 0.198

SFA� saturated f.a.; MUFA�monounsaturated f.a.;PUFA� polyunsaturated f.a.aMann ± Whitney U-test was used to test differences between genders.

Figure 1 Percent of total dietary energy derived from very-long-chainn-3 fatty acids (sum of 20:5, 22:5, 22:6) related to age groups among men(s) and women (d); means and 95% con®dence intervals.

Figure 2 Intake of ®sh (g=10 MJ) related to age groups among men (s)and women (d); means and 95% con®dence intervals.

Intake of very-long-chain n-3 fatty acidsLRK Johansson et al

718

very-long-chain n-3 fatty acids than subjects in the lowestquartile (Table 3). They had 2 ± 3 E% higher fat intake, dueto mono- and polyunsaturated fatty acids. Men and womenin the highest quartile of intake of very-long-chain n-3 fattyacids also had 3-4 fold higher daily intake of retinol andvitamin D, as well as 20 ± 50% higher intake of fruits andvegetables, dietary ®bre and vitamin C. Furthermore, theyused vitamin supplements more frequently. Even if theywere 9 ± 10 y older, they had higher energy intake, and thebody mass index (BMI) was not different from subjects inthe lowest quartile for intake of very-long-chain n-3 fatty

acids. More of both men and women in the highest quartileof intake reported high attention to healthy diet as com-pared to the lowest quartile. The percent of subjects withshort education, blue collar workers and subjects with lessphysical leisure exercise than once weekly, was lower inthe highest as compared to the lowest quartile for intakeamong men, but not among women.

Intake related to social status and lifestyleDietary energy derived from very-long-chain n-3 fattyacids was higher among white vs blue collar workers, aswell as among subjects paying high vs low attention tokeeping a healthy diet (Table 4). Length of education andregular physical leisure exercise was associated with ahigher intake of very-long-chain n-3 fatty acids onlyamong men. Smoking habits were not signi®cantly asso-ciated with dietary intake of very-long-chain n-3 fatty acidsin either gender. The dietary intake of very-long-chain n-3fatty acids did not differ signi®cantly between the six maingeographical regions of Norway or between rural and urbanareas.

Fish intake was signi®cantly lower among subjectsreporting low as compared to high attention paid to healthydiet, among men (73 vs 88 g=d) as well as for women (53 vs62 g=d). Fish intake was also signi®cantly lower amongmale blue vs white collar workers (74 vs 80 g=d), for menwith < 13 vs � 13 y of education (78 vs 82 g=d), andamong males with regular physical leisure activity < 1vs � 1 time=week (71 vs 80 g=d), which was not observedamong women.

For both genders the intake of cod liver oil was sig-ni®cantly lower among blue than white collar workers, forsubjects with short vs long education, among subjects withphysical leisure activity < 1 weekly as compared to � 1

Figure 3 Intake of cod liver oil (g=10 MJ) related to age groups amongmen (s) and women (d); means and 95% con®dence intervals.

Table 3 Dietary intake and characteristics among men and women with low (quartile 1) and high (quartile 4) daily intake ofvery-long-chain n-3 fatty acids; means (s.d.)

Men Women

Quartile 1 Quartile 4 Quartile 1 Quartile 4(n� 379) (n� 378) (n� 406) (n� 406)

Dietary intakeEnergy (MJ=d) 10.3 (3.7) 12.2 (4.6)*** 7.2 (3.0) 8.5 (2.6)***

Fat (E%) 29.7 (6.0) 32.4 (5.7)*** 28.6 (6.1) 32.2 (5.9)**

SFA (E%) 12.2 (3.0) 12.0 (2.8) ns 12.1 (3.0) 12.2 (2.6) nsMUFA (E%) 10.4 (2.2) 11.6 (2.1)*** 9.9 (2.3) 11.4 (2.1)***

PUFA (E%) 5.0 (1.7) 6.2 (1.7)*** 4.6 (1.6) 5.9 (1.8)***

EPA, DPA, DHA, total (E%) 0.08 (0.04) 0.86 (0.47)*** 0.07 (0.04) 0.82 (0.46)***

EPA, DPA, DHA, total (g=d) 0.19 (0.08) 2.57 (1.22)*** 0.13 (0.06) 1.73 (0.84)***

Retinol (mg=d) 0.93 (0.60) 2.59 (2.08)*** 0.73 (0.56) 2.00 (1.24)***

Vitamin D (mg=d) 5.1 (4.4) 23.5 (12.6)*** 5.3 (6.3) 18.5 (11.9)***

Vitamin C (mg=d) 121 (70) 171 (101)*** 124 (76) 162 (83)***

Fibre (g=d) 23 (9) 29 (10)*** 19 (8) 23 (8)***

Fish (g=d) 33 (27) 125 (83)*** 29 (23) 81 (51)***

Cod liver oil (g=d) 0.03 (0.1) 5.0 (4.2)*** 0.02 (0.08) 4.1 (3.9)***

Fruit and vegetables (g=d) 250 (197) 376 (263)*** 273 (196) 367 (237)***

Vitamin supplements (times=week) 1.9 (3.4) 2.8 (4.5)* 3.1 (4.2) 3.9 (4.4)*

Age (y) 37.5 (16.5) 47.8 (15.8)*** 37.8 (17.0) 46.9 (16.6)***

Body mass index (weight=height2) 24.3 (3.2) 24.4 (2.7) ns 23.1 (3.9) 23.4 (3.8) nsEducation � 13 y (%) 14 25*** 19 22 nsBlue collar workers (%) 33 26* 24 21 nsHigh attention to healthy diet (%) 20 40*** 30 49***

Exercise < 1 time=week (%) 36 21*** 24 20 nsNon-smokers (%) 62 62 ns 63 65 ns

*P< 0.05; **P< 0.01; ***P< 0.001 for differences between quartile 1 and 4 within gender. Mann ± Whitney U was used to testdifference in dietary intake and characteristics, and w2 test to test difference in proportions.

Intake of very-long-chain n-3 fatty acidsLRK Johansson et al

719

weekly, and for subjects reporting low vs high attention tohealthy diet. Furthermore, male smokers had a lower intakeof cod liver oil compared to non-smokers.

Discussion

Our analysis started with the assumption that intake ofvery-long-chain n-3 fatty acids would be associated withgender, social status and lifestyle similar to many otherindicators for healthy dietary choices (Blaxter, 1990; PraÈt-taÈlaÈ et al, 1994; Kleemola et al, 1994). However, thepresent study showed that the percent of total dietaryenergy from very-long-chain n-3 fatty acids was slightly,but signi®cantly higher among men than women. Highdietary content of very-long-chain n-3 fatty acids wascorrelated to length of education and regular leisure exer-cise among men, but not among women. Furthermore, theintake of these fatty acids was not signi®cantly correlated tosmoking habits, in contrast to other indicators for a healthydiet (Johansson et al, 1997a; Margetts & Jackson, 1993;JaÈrvinen et al, 1994). Degree of attention paid to keeping ahealthy diet was strongly associated with the intake of very-long-chain n-3 fatty acids in both genders, as expected.Subjects in the highest quartile of intake of these fatty acidsalso had a much higher intake of other healthy foods, suchas fruits and vegetables.

Intake of ®sh and cod liver oil was strong determinantsfor the intake of very-long-chain n-3 fatty acids. Fish intakewas positively correlated to high social status among men,but not among women. The use of cod liver oil onlydiffered slightly between men and women, in contrast touse of vitamin supplements, which were used more oftenamong women than men in the present survey (58 vs 48%)(Johansson et al, 1997a), indicating that women prefer non-fatty in front of fatty supplements. This may have causedthe weaker association between intake of very-long-chainn-3 fatty acids and social status among women than men.

The total intake of linolenic acid and very-long-chain n-3 fatty acids reported in the present survey was 0.9 E%, andabove the recommended minimum level of 0.5 E% (Nordicnutrition recommendations, 1996). The intake of very-long-chain n-3 fatty acids was higher than found among Danes(Bang et al, 1980; Tjùnneland et al, 1993), Swedes (Nydahlet al, 1996) and US health professionals (Ascherio et al,1995), and lower than among Icelanders (Steingrimsdottiret al, 1995). The intake of EPA in the present study

(0.34 g=d) was higher than earlier reported from inlanddistricts of Northern Norway (0.25 g=d), but lower thanreported (0.9 g=d) from coastal districts (Simonsen et al,1988).

The average intake of ®sh in the present study (67 g=d)was higher than in Norwegian household consumptionsurveys and market surveys (National Nutrition Council,1996). Validation studies have shown that the intake of ®shwas higher when reported with the questionnaire than withdietary records (108 vs 75 g=d) (Nes et al, 1992) and 48 hrecall (71 vs 64 g=d) (Solvoll et al, 1993). The ®sh intake inour survey was higher than reported in national dietarysurveys in the Netherlands (approximately 10 g=d) (Hul-shof et al, 1991), Denmark (24 g=d) (Andersen et al, 1996),Sweden (30 g=d) (Becker, 1994), as well as in large pro-spective surveys in the Netherlands (Kromhout et al, 1985)and USA (Daviglus et al, 1997), and secondary preventiontrials in UK (Burr et al, 1989). However, the ®sh intake waslower than reported in national surveys in Iceland (73 g=d)and Japan (approximately 90 g=d) (World Health Organiza-tion, 1990). We have not enough data about the Norwegian®sh consumption to tell if it has changed during the lastdecades (National Nutrition Council, 1996).

The percentage of consumers of cod liver oil was higherin the present study than in earlier Norwegian dietarysurveys (Johansson, 1993) and the average daily intake ofcod liver oil was higher than the total Norwegian produc-tion of cod liver oil (approximately 0.3 g=d=inhabitant) inthe 1980s and early 1990s (Johansson, 1993). Marketing ofmarine oils has been extensive and it is likely that theintake of cod liver oil has increased in Norway during the1990s. However, the average intake of cod liver oil in ourstudy was less than half reported by Icelanders (3 g=d)(Steingrimsdottir et al, 1995). It was also much lower thanfound among Norwegian families during the war 1942 ±1945 when the intake of cod liver oil and ®sh was 4 g=dand 285 g=d, respectively (Strùm, 1948).

Conclusions

In summary, the range of energy derived from very-long-chain n-3 fatty acids was very large in the populationstudied. Men had slightly higher intake of these fattyacids compared to women. High intake of very-long-chain n-3 fatty acids was associated with high socialstatus and regular physical leisure exercise among men.

Table 4 Dietary energy from very-long-chain n-3 fatty acids (E%) related to social and lifestyle variables amongmen and women; means (s.d.)

Number Men Pa Number Women Pa

Occupationblue collar worker 404 0.32 (0.33) 370 0.30 (0.30)white collar worker 457 0.41 (0.41) < 0.001 459 0.36 (0.39) 0.048

Education< 13 y 1139 0.37 (0.39) 1224 0.35 (0.37)� 13 y 307 0.43 (0.39) < 0.001 316 0.37 (0.40) 0.730

Attention to healthy dietvery low, low and medium 1086 0.32 (0.34) 993 0.30 (0.31)high and very high 428 0.50 (0.48) < 0.001 630 0.43 (0.43) < 0.001

Physical leisure exercise< 1 time=week 441 0.31 (0.31) 367 0.32 (0.34)� 1 time=week 1067 0.40 (0.41) < 0.001 1247 0.36 (0.38) 0.118

Smoking habitsnon-smoker 926 0.39 (0.40) 1040 0.36 (0.38)smoker 591 0.36 (0.37) 0.275 587 0.34 (0.37) 0.587

aMann ± Whitney U-test was used to evaluate differences between subgroups within genders.

Intake of very-long-chain n-3 fatty acidsLRK Johansson et al

720

In both genders intake of very-long-chain n-3 fatty acidswas correlated to other indicators for a healthy diet, exceptnon-smoking.

Acknowledgements Ð The study was carried out in cooperation with the

Norwegian Food Control Authority. The contributions of Christina Berg-

sten, Bodil Blaker, Elin B Lùken and Gunnar AÊ mlid are highly appre-

ciated.

References

Andersen NL, Fagt S, Groth MV, Hartkopp HB, Mùller A, Ovesen L &Warming DL (1996). Danish Dietary Habits 1995 (English summary).National Food Authority, Report No. 235. Quickly Tryk A=S, Copen-hagen.

Ascherio A, Rimm EB, Stampfer MJ, Giovannucci EL & Willett WC(1995). Dietary intake of marine n-3 fatty acids, ®sh intake, and therisk of coronary heart disease among men. N. Engl. J. Med. 332,977 ± 982.

Bang HO, Dyerberg J & Sinclair HM (1980). The composition of theEskimo food in north western Greenland. Am. J. Clin. Nutr. 33,2657 ± 2661.

Becker W (1994). Dietary Habits and Nutrient Intake in Sweden 1989.Methods and Results (English summary). TK i Uppsala: National FoodAuthority.

Blaker B & Aarsland M (1989). Household Measures and Weights ofFoods (in Norwegian). Engers Boktrykkeri A=S. Otta: National Asso-ciation for Nutrition and Health.

Blaxter M (1990). Health and Lifestyles. London: Tavistock=Routledge.Brude IR, Drevon CA, Hjermann I, Selje¯ot I, Lund-Katz S, Saarem K,

Sandstad B, Solvoll K, Halvorsen B, Arnesen H & Nenseter MS (1997).Peroxidation of LDL from combined-hyperlipidemic male smokerssupplied with omega-3 fatty acids and antioxidants. Arterioscler.Thromb. Vasc. Biol. 17, 2576 ± 2588.

Burr ML, Fehily AM, Gilbert JF, Rogers S, Holliday RM, Sweetnam PM,Elwood PC & Deadman NM (1989). Effects of changes in fat, ®sh, and®bre intakes on death and myocardial infarction: Diet and reinfarctiontrial (DART). Lancet 334, 757 ± 761.

Bùnaa KH, Bjerve KS, Straume B, Gram IT & Thelle D (1990). Effect ofeicosapentaenoic and docosahexaenoic acids on blood pressure inhypertension. A population-based intervention trial from the TromsùStudy. N. Engl. J. Med. 332, 795 ± 801.

Central Bureau of Statistics (1984). Standard Classi®cation of Socio-Economic Status. (SNS no. 5). Oslo, Kongsvinger.

Daviglus ML, Stamler J, Orencia AJ, Dyer AR, Liu K, Greenland P, WalshMK, Morris D & Shekelle RB. (1997). Fish consumption and the30-year risk of fatal myocardial infarction. N. Engl. J. Med. 336,1046 ± 1053.

de Lorgeril M, Renaud S, Mamelle N, Salen P, Martin J-L, Monjaud I,Guidollet J, Touboul P & Delaye J (1994). Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heartdisease. Lancet 343, 1454 ± 1459.

Dolecek TA (1992). Epidemiological evidence of relationships betweendietary polyunsaturated fatty acids and mortality in the multiple riskfactor intervention trial. Proc. Soc. Exp. Biol. Med. 200, 177 ± 182.

Drevon CA (1992). Marine oils and their effects. Nutr. Rev. 50, 38 ± 45.Frost Andersen L, Solvoll K & Drevon CA (1996). Very long-chain n-3

fatty acids as biomarkers for intake of ®sh and n-3 fatty acid concen-trates. Am. J. Clin. Nutr. 64, 305 ± 311.

Hjartaker A, Lund E & Bjerve KS (1997). Serum phospholipid fatty acidcomposition and habitual intake of marine foods registered by a semi-quantitative food frequency questionnaire. Eur. J. Clin. Nutr. 51,736 ± 742.

Hulshof KFAM, LoÈwik MHR, Kok FJ, Wedel M, Brants HAM, HermusRJJ & ten Hoor F (1991). Diet and other life-style factors in high andlow socio-economic groups (Dutch Nutrition Surveillance System).Eur. J. Clin. Nutr. 45, 441 ± 450.

JaÈrvinen R, Knekt P, SeppaÈnen R, Reunanen A, HelioÈvaara M, Maatela J& Aromaa A (1994). Antioxidant vitamins in the diet: relationships withother personal characteristics in Finland. J. Epidem. Com. Health. 48,549 ± 554.

Johansson L (1993). Kostholdets betydning for dùdeligheten av hjertein-farkt i Norge de siste 100aÊr (In Norwegian). Statens ernñringsraÊd ogAvdeling for kostholdsforskning, Universitetet i Oslo. Lobo Gra®sk as,Oslo.

Johansson L, Solvoll K, Bjùrneboe G-E Aa & Drevon CA (1997a). Dietaryhabits among Norwegian men and women. Scand. J. Nutr. 41, 63 ± 70.

Johansson L, Solvoll K, Opdahl S, Bjùrneboe G-E Aa & Drevon CA(1997b). Response rates with different distribution methods and reward,and reproducibility of a quantitative food frequency questionnaire. Eur.J. Clin. Nutr. 51, 346 ± 353.

Kleemola P, Virtanen M & Pietinen P (1994). The 1992 Dietary Survey ofFinnish Adults. Publications of the National Public Health Institute,B2=1994. National Public Health Institute, Helsinki.

Kromhout D, Bosschieter EB & de Lezenne Coulander C (1985). Theinverse Relation between ®sh Consumption and 20-Year Mortality fromCoronary Heart Disease. N. Engl. J. Med. 312, 1205 ± 1209.

Margetts BM & Jackson AA (1993). Interactions between people's dietand their smoking habits: the dietary and nutritional survey of Britishadults. BMJ 307, 1381 ± 1384.

McKeigue P (1994). Diets for secondary prevention of coronary heartdisease: can linolenic acid substitute for oily ®sh? Lancet 343, 1445.

Morris MC, Manson JE, Rosner B, Buring JE, Willett WC & HennekensCH (1995). Fish consumption and cardiovascular disease in the physi-cians' health study: a prospective study. Am. J. Epidemiol. 142,166 ± 175.

National Nutrition Council (1996). Trends in the Norwegian Diet(Norwegian). Oslo.

Nes M, Frost Andersen L, Solvoll K, Sandstad B, Hustvedt BE, Lùvù A &Drevon CA (1992). Accuracy of a quantitative food frequency ques-tionnaire applied in elderly Norwegian women. Eur. J. Clin. Nutr. 42,809 ± 821.

Nestel PJ (1987). Polyunsaturated fatty acids (n-3, n-6). Am. J. Clin. Nutr.45, 1161 ± 1167.

Nordic Nutrition Recommendations (1996). Scand. J. Nutr. 40, 161 ± 165.Nydahl M, Gustafsson I-B, Mohsen R & Vessby B (1996). The food and

nutrient intake of Swedish non-smokers and smokers. Scand. J. Nutr.40, 64 ± 69.

Pietinen P, Ascherio A, Korhonen P, Hartman AM, Willett W C, AlbanesD & Virtamo J (1997). Intake of fatty acids and risk of coronary heartdisease in a cohort of Finnish men. Am. J. Epidemiol. 145, 876 ± 887.

PraÈttaÈlaÈ R, Karisto A & Berg M-A (1994). Consistency and variation inunhealthy behaviour among Finnish men, 1982 ± 1990. Soc. Sci. Med.39, 115 ± 122.

Simonsen T, Nordùy A, Sjunneskog C & Lyngmo V (1988). The effect ofcod liver oil in two populations with low and high intake of dietary ®sh.Acta. Med. Scand. 223, 491 ± 498.

Solvoll K, Lund-Larsen K, Sùyland E, Sandstad B & Drevon CA (1993). Aquantitative food frequency questionnaire evaluated in a group ofdermatologic outpatients. Scand. J. Nutr. 37, 150 ± 155.

SPSS for Windows (1996). Release 7.5. SPSS Inc. 444N. MichiganAvenue, Chicago, IL, 60611.

Statistics Norway (1995). Statistical Yearbook 1995. Of®cial Statistics ofNorway. C247. Oslo, Kongsvinger.

Steingrimsdottir L, Sigurdsson Jr G & Sigurdsson G (1995). Nutrition andserum lipids in Iceland. Scand. J. Nutr. 39, 138 ± 141.

Strùm A (1948). Examination of the diet of Norwegian families during thewar 1942 ± 1945. Act. Med. Scand. (Suppl).

Sùyland E, Funk J, Rajka G, Sandberg M, Thune P, Rustad L, Helland S,Middelfart K, Odu S, Falk ES, Solvoll K, Bjùrneboe G-E & Drevon CA.(1993). Effect of dietary supplementation with very-long-chain n-3 fattyacids in patients with psoriasis. N. Engl. J. Med. 328, 1812 ± 1816.

Sùyland E, Funk J, Rajka G, Sandberg M, Thune P, Rustad L, Helland S,Middelfart K, Odu S, Falk ES, Solvoll K, Bjùrneboe G-E & Drevon CA.(1994). Dietary supplementation with very long-chain n-3 fatty acids inpatients with atopic dermatitis. A double-blind, multicentre study. Br. J.Dermatol. 130, 757 ± 764.

Tjùnneland A, Overvad K, Thorling E & Ewertz M (1993). Adipose tissuefatty acids as biomarkers of dietary exposure in Danish men andwomen. Am. J. Clin. Nutr. 57, 629 ± 633.

Vollset SE, Heuch I & Bjelke E (1985). Fish consumption and mortalityfrom coronary heart disease. N. Engl. J. Med. 313, 820 ± 821.

World Health Organization (1990). Diet, Nutrition and the Prevention ofChronic Diseases. Report of a WHO Study Group. Technical ReportSeries 797. Geneva: WHO.

Intake of very-long-chain n-3 fatty acidsLRK Johansson et al

721