Fish consumption and polyunsaturated fatty acids in relation to psychological distress

Abstract

Background It has been suggested that high fish consumption improves mental well-being. The aim of this study was to assess whether high fish consumption or omega-3 polyunsaturated fatty acid (PUFA) intake was associated with reduced self-reported psychological distress.

Methods We used three cross-sectional data sets, the nationwide Health 2000 Survey (n = 5840), the Fishermen Study on Finnish fishermen and their family members (n = 1282) and the Finntwin16 Study on young adults (n = 4986). Data were based on self-administered questionnaires, interviews, health examinations and blood samples. Psychological distress was measured using the 12-item and 21-item General Health Questionnaires (GHQs). Fish consumption was measured by a food frequency questionnaire (FFQ, g/day) and independent frequency questions (times/month). Dietary intake (g/day) and serum concentrations (% from fatty acids) of PUFAs were determined. Relationships were analysed using regression analysis.

Results Regardless of the measure, fish consumption and omega-3 PUFA dietary intake were not associated with distress in any of the data sets. In contrast to the hypothesis, high serum docosahexaenoic acid was associated with high distress in the Fisherman Study men. Some non-linear associations were detected between serum omega-3 PUFAs or fish consumption (times/month) and distress. In the Fishermen Study, the associations were modified by alcohol consumption, smoking and physical activity.

Conclusions Our results do not support the hypothesis that fish consumption or omega-3 PUFA intake are associated with reduced psychological distress in the general population or in a population with high fish consumption.

Epidemiology, mental well-being, fish consumption, polyunsaturated fatty acids (PUFAs), National Survey

Introduction

It has been suggested that high fish consumption and high intake of omega-3 polyunsaturated fatty acids (PUFAs) promote mental well-being. Among the omega-3 PUFAs, eicosapentaenoic acid (EPA, 20:5n–3) and docosahexaenoic acid (DHA, 22:6n–3) are thought to be the most beneficial.^1,2 Several cross-national and country-specific epidemiological studies suggest that high fish consumption is correlated particularly with lower rates of depression or fewer depressive symptoms.^3–7 Low tissue concentrations of EPA and DHA—either absolute or relative to arachidonic acid (AA, 20:4n–6)—analysed from serum/plasma, red blood cells or adipose tissue appear to be associated with a greater severity of symptoms in both affective and psychotic disorders.⁸ However, controversial findings exist.^9–16 A recent review concluded that the current evidence of the beneficial effects of omega-3 PUFAs on mood is limited and highly inconsistent, both in terms of study methodology and findings.¹⁷

The General Health Questionnaire (GHQ) is a screening device for identifying minor psychiatric disorders or the presence of psychological distress in non-psychiatric populations.¹⁸ The rationale behind the instrument is that all psychiatric disorders share a common underlying element of psychological distress. The GHQ is a very common measure of mental well-being but has been little utilized in studies on the health effects of fish consumption and omega-3 PUFA intake. In a clinical trial among smokers, nutritional supplements containing fish oils (mainly DHA) alleviated psychological distress measured by the 30-item GHQ.¹⁹ To our knowledge, there are no previous observational studies assessing fish consumption or omega-3 PUFA intake in relation to psychological distress measured by the GHQ.

To investigate the hypothesized association between fish consumption and mental well-being, we utilized three distinct cross-sectional data sets: the Health 2000 Survey represented the general population of Finland; the Nutrition, Environment and Health Study (the Fishermen Study) represented a population with high fish consumption; and the Finntwin16 Study represented young adults. The aim was to assess whether high fish consumption and omega-3 PUFA intake was associated with reduced self-reported psychological distress, as measured by the GHQ.

Methods

Study design

The nationally representative Health 2000 Survey was carried out in 2000–01.²⁰ The sample comprised of persons aged ≥30 years (n = 8208). Data were gathered by structured interviews, health examinations and self-administered questionnaires. A sub-sample of participants aged 45–74 years and living near five central university hospitals (n = 1526), were invited to participate in further examinations, including blood sampling in 2001–02.

The Fishermen Study population consisted of all Finnish maritime and freshwater area fishermen, their wives and other family members identified from the Professional Fishermen Register and the national Population Register Centre.²¹ For a postal questionnaire study, a national random sample of 4378 fishermen, and their wives and other family members, was drawn from the registers. From 1427 respondents, 309 persons living near Helsinki and Turku attended a sub-study that included a health examination and blood sampling in 2004–05.

The Finntwin16 Study is a nation-wide longitudinal cohort study on health behaviour among young adult twins and their families in Finland. Virtually all twin births from 1975 through 1979 were identified from the national Population Register Centre of Finland.²² A questionnaire was mailed semi-annually between 2000 and 2002 to each of the five birth cohorts of twin pairs. Of those, 5043 individuals (85%) responded.²³

Participants with mean energy intakes >6999 or <600 kcal were excluded in the Health 2000 Survey. Furthermore, participants with missing data on fish consumption or omega-3 PUFA intake, or with more than two items missing in the GHQ (n = 140 in the Health 2000 Survey, n = 17 in the Fishermen Study and n = 16 in the Finntwin16 Study), were excluded. The final numbers of the study participants were n = 5840 in the Health 2000 Survey, n = 1288 in the Health 2000 sub-study, n = 1410 in the Fishermen Study, n = 308 in the Fishermen sub-study and n = 4986 in the Finntwin16 Study.

Main variables

The study outcome was psychological distress measured by the 12-item GHQ in the Health 2000 Survey and the Fishermen Study, and the 21-item GHQ (the standard 20-item GHQ with one additional item) in the Finntwin16 Study.²⁴ To achieve more statistical power,²⁵ we retained the Likert scoring (0-1-2-3) corresponding to the four response options of each item: ‘not at all’, ‘same as usual’, ‘rather more than usual’ and ‘much more than usual’. A high score in any question indicated greater distress. The range was 0–36 for the GHQ-12 and 0–63 for the GHQ-21.

In the case of one or two missing values in the GHQ items, the missing values were replaced by the mean value of the remaining GHQ items. In the Health 2000 Survey, 2.3% of the participants had one missing item and 0.4% had two missing items. The respective figures were 1.3 and 0.5% in the Fishermen Study, and 3.3 and 0.3% in the Finntwin16 Study.

The participants of the Health 2000 Survey (n = 5840) and the Fishermen sub-study (n = 308) completed a validated^26,27 food frequency questionnaire (FFQ) designed to assess the whole diet over the previous 12 months.²⁶ The FFQ consisted of 128 food items grouped under 12 subheadings, one of them being fish dishes. Nine response options ranged from ‘never or seldom’ to ‘six or more times per day’. The portion sizes were fixed and, if possible, specified using natural units (e.g. cups of coffee). Fish consumption was converted into grams per day by multiplying the consumption frequency of different fish dishes by fixed portion sizes. Foods were broken down into their components as well as nutrients, including PUFAs, alcohol and energy, using the Finnish Food Composition Database (http://www.fineli.fi/index.php).

In the Fishermen Study, the respondents were also asked about the consumption frequency of different fish dishes and fish species. Six response categories ranged from ‘never’ to ‘almost every day’. From the frequencies, the sums of all fish dishes, fatty fish species and lean fish species were calculated as times per month. The sum of all fish dishes was shown to measure fish consumption equally well when compared with the FFQ on the whole diet.²⁸

In the Finntwin16 Study, the respondents were asked about fish consumption frequency as part of a brief FFQ.²⁹ The response options ranged from ‘never’ to ‘several times a day’.

Fasting blood samples were collected to analyse serum concentrations of fatty acids in the Health 2000 sub-study (n = 1282) and the Fishermen sub-study (n = 300). Serum total fatty acid composition was analysed using gas chromatograph (capillary column, flame ionization detector).³⁰ The analyses were performed at the National Public Health Institute’s testing laboratory of clinical chemistry in Turku,²⁸ and the concentrations were expressed as a proportion from all serum fatty acids.

Evidence suggests that both deficiency in omega-3 PUFA intake and imbalance in the omega-3/omega-6 PUFA ratio may be responsible for depression or depressive symptoms.^17,31 The potential mechanism is related to an hypothesized anti-inflammatory origin of psychiatric diseases. A low omega-3/omega-6 PUFA ratio leads to a high proportion of AA and a small proportion of EPA and DHA in cell membranes. This in turn results in an overproduction of eicosanoids derived from omega-3 PUFAs such as prostacyclin, thromboxane, leuktotrienes and pro-inflammatory mediators (cytokines and leuktotrienes).³² The DHA/AA and EPA/AA ratios are indirect measures of the balance of omega-3/omega-6 PUFAs in the diet and represent comparisons of omega-3 and omega-6 PUFAs at different stages of fatty acid synthesis. Therefore, dietary omega-3/omega-6 PUFA, DHA/AA and EPA/AA ratios were calculated.

Information on demographic and economic background, self-reported health, medication and health behaviour was obtained from an interview in the Health 2000 Survey and from self-administered questionnaires in the Fishermen Study and in the Finntwin16 Study (categorized as shown in Table 1). In the Health 2000 Survey and the Fishermen sub-study, weight and height were mainly measured during the health examination. Self-reported height and weight were used to calculate body mass index (BMI; kg/m²) in the Fishermen Study and the Finntwin16 Study. The latter has been validated in a sub-sample.^33,34

Table 1

Distribution of participants in the Health 2000 Survey, the Fishermen Study and the Finntwin16 Study

	The Health 2000 Survey (in 2000–01)		The Fishermen Study (in 2004)		The Finntwin16 Study (in 2000–02)
	Men n = 2608	Women n = 3232	Men n = 610	Women n = 800	Men n = 2303	Women n = 2683
Age (years), mean (SE)	51 (0.3)	53 (0.3)	47 (0.5)	46 (0.5)	25 (0.02)	24 (0.02)
Level of education, %
Basic	36	38	40	27	5	4
Intermediate	39	28	43	36	50	45
Higher	25	34	17	37	45	51
Missing (n)	(8)	(10)	(6)	(6)
Marital status, %
Married or cohabiting	77	67	77	86	42	51
Divorced, separated or single parent	8	11	5	3	<1	3
Widow	3	12	<1	3	0	0
Single	12	10	18	8	58	46
Missing (n)	(9)	(5)	(3)	(3)	(3)	(3)
Physical activity, %
Sufficient	30	34	29	33	51	42
Intermediate	30	30	40	39	26	41
Sedentary	40	36	31	28	23	17
Missing (n)	(8)	(53)			(5)	(4)
BMI (kg/m2), mean (SE)	27 (0.08)	27 (0.09)	27a (0.2)	26a (0.2)	24 (0.1)	22 (0.1)
Missing (n)	(1)	(10)	(1)	(17)	(16)	(12)
Smoking history, %
Daily	28	18	23	13	32	24
Occasional or former	36	18	35	21	28	28
Never smoker	36	64	42	65	40	48
Missing (n)	(18)	(14)	(12)	(12)		(1)
Alcohol-induced intoxication, %
At least once a week	–	–	14	2	22	8
At least once a month	–	–	28	11	43	34
Less frequently	–	–	43	49	26	45
Never	–	–	14	38	9	13
Alcohol (ethanol, g/day), mean (SE)	8 (0.3)	3 (0.1)	12b (1.3)	4b (0.4)	13 (0.4)	5 (0.2)
Energy (MJ/day), mean (SE)	10 (0.08)	9 (0.06)	10b (0.3)	9b (0.2)	–	–
Missing (n)	(1)	(10)	(1)	(17)
Fish oil supplement user, %	1	3	5	8	–	–
Occurrence of
Severe illness, %c	17	20	11	11	11	12
Missing (n)	(13)	(12)	(44)	(40)	(22)	(27)
Back pain or illness, %d	33	31	20	17	10	7
Missing (n)	(9)	(9)	(31)	(37)	(14)	(14)
Bronchial asthma, %	7	10	3	4	7	7
Missing (n)	(7)	(6)	(35)	(41)	(3)	(4)
Medication for depression or psychiatric disorders, %e	4	7	9	12	16	24
Missing (n)	(300)	(118)	(7)	(6)	(4)	(8)
Psychological distress (GHQf Likert score), mean (SE)	11/36 = 0.3 (0.1)	12/36 = 0.3 (0.1)	10/36 = 0.3 (0.2)	11/36 = 0.3 (0.2)	17/63 = 0.3 (0.2)	20/63 = 0.3 (0.2)

SE: standard error.

^aSelf-reported.

^bAvailable only for the sub-study (n = 308).

^cCancer, myocardial infarction, cerebral stroke, angina pectoris, cardiac insufficiency, diabetes or rheumatoid arthritis. Cancer, myocardial infarction and cerebral stroke ever diagnosed by a physician in both studies. Other diseases ever diagnosed by a physician in the Health 2000 Survey, diagnosed or treated by a physician during the previous 12 months in the Fishermen Study, and any self-reported chronic condition in the Finntwin16 Study.

^dPercentage of individuals with 10 or more occurrences of back pain necessitating absence from work/study in the Fintwin16 study.

^eAny medication in the Finntwin16 Study.

^f12-item General Health Questionnaire (GHQ) in the Health 2000 Survey and in the Fishermen Study, 21-item GHQ in the Finntwin16 Study.

Statistical methods

In the Health 2000 Survey, weights were used for handling correlated data with unequal sampling probabilities and for correcting the effects of over-sampling people aged 80 years. Nutrient intakes were adjusted for total energy intake using the residual method³⁵ both in the Health 2000 Survey and in the Fishermen Study.

Multivariable linear regression models were used to study the linear associations between fish consumption or PUFAs and distress. The GHQ-12 scores were log-transformed to fulfil model assumptions. The distribution of GHQ-21 scores was satisfactorily normal and therefore scores were not transformed. Missing values for categorized covariates were included as an additional category. For continuous covariates, missing values were replaced by mean values in the Health 2000 Survey and the Fishermen Study. In the Finntwin16 Study, regression models for fish consumption frequencies were run using three dummy variables. To adjust for the possible lack of statistical independence of twins sampled as twin pairs, robust estimators of variance were used to obtain correct standard errors and P-values using the cluster option in the linear regression.³⁶ To allow for possible non-linearity, regression models that included quadratic terms for the FFQ fish consumption and PUFA intake were conducted as well.

Possible effect modification was studied by adding product terms for quartiles of FFQ fish consumption and PUFA intake with educational level, smoking and physical activity categorized as shown in Table 1, and alcohol consumption (ethanol, g/day) divided into tertiles. The product terms were added one by one into the fully adjusted models.

Results

Except for smoking and alcohol consumption, the Health 2000 Survey and the Fishermen Study participants were quite similar regarding the variables shown in Table 1. Moreover, the demographic characteristics of the participants in both the sub-studies were similar to the main studies (data not shown). The Finntwin16 participants were much younger, better educated and more often living alone or with their parents when compared with the other two studies. They also used more alcohol and smoked more on a non-daily basis, but were physically more active.

Psychological distress as measured by the GHQ was at the same level in each data set (Table 1). Fish consumption and both the intake and serum concentrations of omega-3 PUFAs were high among the Fishermen Study participants (Table 2). Differences between the Health 2000 Survey and the Fishermen Study were the largest in serum DHA; the lowest quartile of the Fishermen sub-study men corresponded to the third quartile of the male general population. In the Finntwin16 Study, 1.0% of men reported eating fish daily, 27% a few times a week, 65% a few times a month and 7.6% never. Of women, the respective figures were 0.8, 27, 64 and 7.8%.

Table 2

Fish consumption and the intake of PUFAs in the Health 2000 Survey and the Fishermen Study

	The Health 2000 Survey				The Fishermen Study
	Men		Women		Men		Women
	Mean	(95% CI)	Mean	(95% CI)	Mean	(95% CI)	Mean	(95% CI)
Frequency questions on fish consumption (times/month)			(n = 610)		(n = 800)
Fish	–		–		13	12–14	9.5	9.0–10
Fatty fish	–		–		5.3	4.8–5.8	4.6	4.2–4.9
Lean fish	–		–		5.5	4.9–6.1	3.8	3.4–4.1
FFQ fish consumption (g/day)	(n = 2608)		(n = 3232)		(n = 142)		(n = 166)
Fish	38	36–39	37	36–39	67	57–77	51	46–56
Fish products	8.0	7.6–8.5	6.9	6.5–7.3	12	9.7–14	7.9	6.5–9.2
Fish and fish products	46	44–48	44	43–46	79	68–89	59	54–65
FFQ fatty acid intake (g/day)	(n = 2608)		(n = 3232)		(n = 142)		(n = 166)
Omega-3 PUFAs	2.7	2.7–2.8	2.8	2.8–2.8	3.4	3.2–3.6	3.3	3.1–3.4
EPA	0.2	0.2–0.2	0.2	0.2–0.2	0.3	0.2–0.3	0.2	0.2–0.2
DHA	0.5	0.5–0.5	0.5	0.5–0.5	0.7	0.6–0.8	0.6	0.5–0.6
Alpha-linolenic acid	1.8	1.8–1.8	1.9	1.9–1.9	2.1	2.0–2.1	2.2	2.1–2.3
Omega-6 PUFAs	12	12–12	12	12–12	10	9.9–11	11	10–11
Linoleic acid	11	11–12	12	11–12	8.7	8.4–9.0	8.9	8.6–9.7
Gammalinolenic acid	0.1	0.1–0.1	0.1	0.1–0.1	0.04	0.04–0.05	0.1	0.04–0.1
AA	0.1	0.1–0.1	0.1	0.1–0.1	0.1	0.1–0.1	0.1	0.1–0.1
EPA/AA ratio	1.6	1.6–1.7	1.8	1.8–1.9	2.1	1.9–2.3	2.1	1.9–2.2
DHA/AA ratio	4.3	4.2–4.4	4.9	4.7–5.0	5.3	4.8–5.9	5.4	5.0–5.8
Omega-3/omega-6 PUFAs ratio	0.2	0.2–0.2	0.2	0.2–0.2	0.3	0.3–0.3	0.3	0.3–0.3
Serum fatty acid concentration (% from all fatty acids)	(n = 572)		(n = 710)		(n = 139)		(n = 161)
EPA + DHA + DPA	3.9	3.8–4.0	4.0	3.8–4.1	6.9	6.4–7.5	6.7	6.3–7.1
EPA	1.2	1.2–1.3	1.2	1.2–1.3	2.2	1.9–2.5	1.8	1.5–2.0
DHA	2.2	2.1–2.2	2.2	2.2–2.3	4.0	3.7–4.3	4.3	4.0–4.6
DPA	0.5	0.5–0.5	0.5	0.5–0.5	0.7	0.6–0.7	0.6	0.6–0.7
AA	4.4	4.4–4.5	4.4	4.4–4.5	6.0	5.8–6.2	6.0	5.8–6.1

95% CI: 95% confidence interval; FFQ: food frequency questionnaire; EPA: eicosapentaenoic acid; DPA: docosapentaenoic acid; AA: arachidonic acid.

Except for age, education and BMI in the young adults, the GHQ scores were associated with all the covariates in each of the data sets. Furthermore, in all the data sets, health behaviour in terms of smoking and physical activity was more favourable in participants with high fish consumption when compared with those of low fish consumption. In the Health 2000 Survey and the Fishermen Study men, higher fish consumption was associated with higher alcohol (ethanol, g/day) consumption.

Regardless of the measure, fish consumption or omega-3 PUFA dietary intake was not associated with distress in any of the data sets. In contrast to the hypothesis, high serum DHA was associated with high distress scores in the Fisherman Study men (β = 0.02) implying an 2.0% increase in the log-transformed distress scores for every 10% increase in serum DHA (Table 3).

Table 3

Fish consumption and the intake of omega-3 PUFAs^a in relation to psychological distress measured by the 12-item GHQ in linear regression models in the Health 2000 Survey and the Fishermen Study

	Psychological distress (log-transformed GHQ-12 Likert scores)
	The Health 2000 Survey				The Fishermen Study
	Men		Women		Men		Women
	β	P-value	β	P-value	β	P-value	β	P-value
FFQ fish and fish products consumption (10 g/day)
	0.0004	0.67	−0.001	0.67	0.001	0.80	−0.003	0.51
FFQ fatty acid intake (100 mg/day)
EPA	−0.002	0.48	0.0001	0.95	−0.002	0.78	−0.01	0.29
DHA	−0.0001	0.66	0.0002	0.80	−0.001	0.74	−0.01	0.34
EPA/AA ratio	−0.002	0.68	−0.001	0.71	−0.003	0.78	−0.003	0.85
DHA/AA ratio	−0.0002	0.89	−0.0002	0.86	−0.001	0.76	−0.001	0.87
Omega-3/omega-6 PUFA ratio	−0.03	0.61	−0.03	0.55	−0.09	0.50	−0.17	0.54
Serum fatty acid concentration (% from all fatty acids)
EPA	−0.01	0.52	−0.003	0.73	0.001	0.46	0.01	0.49
DHA	0.003	0.72	−0.003	0.72	0.02	0.01	0.001	0.95

^aAdjusted for age, level of education, marital status, smoking history, physical activity, total energy intake (except for the models including serum concentrations of fatty acids), alcohol intake, alcohol-induced intoxication (in the Fishermen Study), BMI, medication for depression or psychiatric disorders, occurrence of physician-diagnosed severe illnesses (i.e. cancer, myocardial infarction, cerebral stroke, diabetes or rheumatoid arthritis), back pain or illness, bronchial asthma and regular use of fish oil supplements.

PUFA: polyunsaturated fatty acid; GHQ: General Health Questionnaire; FFQ: food frequency questionnaire; EPA: eicosapentaenoic acid; DHA: docosahexaenoic acid; AA: arachidonic acid.

In the Health 2000 Survey men, a weak non-linear relationship between serum EPA (P = 0.05) and distress was detected in the regression model that included a quadratic term. Non-linear relationships were also detected between serum DHA (P = 0.09) and distress in the Fishermen Study women, and between total (P = 0.02) or lean fish (P = 0.03) consumption (times/month) and distress in the Fishermen Study men.

Due to statistically significant interactions (P < 0.1), stratified regression analyses were performed by tertiles of alcohol consumption, smoking and physical activity. In the Health 2000 Survey, no significant associations were observed between fish consumption (10 g/day) or PUFA intake (100 mg/day) and distress. In the Fishermen Study, the unexpected positive association between high serum DHA and high distress scores remained in men who were in the lowest alcohol consumption tertile (P = 0.003) or who were occasional smokers (P = 0.06). This association disappeared when stratified by physical activity. Instead, high fish consumption (P = 0.04) and high intake of EPA (P = 0.03) and DHA (P = 0.02) were associated with reduced distress scores in the Fishermen Study men who reported having a sedentary lifestyle. In addition, high fish consumption (P = 0.002) and high serum DHA (P = 0.09) were associated with reduced distress scores in the Fishermen Study women who smoked daily.

Discussion

Our results do not support the hypothesis that high fish consumption or omega-3 PUFA intake is associated with reduced psychological distress. The associations, if anything, could be non-linear. An effect modification by health behaviour is also suggested. The beneficial health effects of high fish consumption or omega-3 PUFA intake may manifest only in individuals with detrimental health behaviour such as smoking or a sedentary lifestyle. However, results concerning the observed non-linear associations and an effect modification were inconsistent across the sub-groups. Thus, our main finding is that high fish consumption does not protect from psychological distress. Regardless of the measures of fish consumption and omega-3 PUFA intake, the main finding was consistent in all the three cross-sectional data sets.

The use of three separate data sets and several methods to assess fish consumption and omega-3 PUFA intake increased the validity of the results. In addition, the comprehensive data sets made it possible to evaluate the effects of several confounding and modifying factors. However, it is possible that we were unable to see potential associations due to a cross-sectional setting. The possible associations could be established in longitudinal settings. The nationally representative Health 2000 Survey is a clear strength of our study. The response rates were high in the Health 2000 Survey and the Finntwin16 Study. The response rate was low in the Fishermen Study questionnaire, reducing generalizability of the results. However, these data include additional measures of fish consumption. The professional fishermen and their families represented a population with high habitual fish consumption wherein the health effects of fish should most likely to be manifested. Conversely, fish consumption among the young adults of the Finntwin16 study was low, with less than one-third reporting eating fish a few times a week. Regarding fish consumption, we thus had three different kinds of data. The nationally representative Health 2000 Survey enabled us to investigate the whole range of fish consumption in the general population, e.g. including the extreme ends. Therefore somewhat higher mean fish consumption in Finland³⁷ compared with some other countries in Europe³⁸ did not cause a problem. A limitation of the present study is that both the Health 2000 and the Fishermen sub-studies were based on convenience samples. Results based on these data cannot be generalized since bias introduced by subject participation is possible. In addition, the number of participants in the Fishermen sub-study was low.

In the present study, no hypothesized associations were observed between fish consumption or omega-3 PUFA intake and psychological distress. One reason might be the fact that the study outcome was recent psychological distress instead of more severe mental disorders. The GHQ was developed as a screening tool to detect individuals who are likely to have or to be at risk of developing psychiatric disorders.³⁹ More specifically, it assesses changes in an individual’s ability to carry out daily functions and the emergence of a new psychological disorder rather than life-long personality characteristics. Previous research has mainly focused on more severe psychiatric disorders such as depressive symptoms, depression or even schizophrenia.⁴⁰ In agreement with our findings, a recent study suggested that omega-3 PUFAs may not have a role in the aetiology of minor depression.¹¹ However, according to two clinical trials, supplemental EPA and DHA have reduced stress measured by the GHQ in smokers¹⁹ and measured by the Perceived Stress Scale and the Daily Hassles and Uplifts Scale in a patient population.⁴¹ It is noteworthy that in the first trial, the number of participants was very small (n = 7) and in the second trial, the observed benefits were additional to psychiatric care.

We used the 128-item FFQ from two data sets to measure fish consumption and dietary fatty acid intake and were able to take into account total energy intake in the analyses. This was not done in the majority of the previous studies that showed associations between high fish consumption or dietary intake of omega-3 PUFAs and lower rates of depression or fewer depressive symptoms. In only one of these studies on elderly men were fish consumption and omega-3 PUFA intake adjusted for total energy intake.⁴ On the contrary, total energy intake was used as a covariate in three out of six questionnaire-based studies reporting no linear associations between high fish consumption and lower rates depression or fewer depressive symptoms.^10,12–16 Hence, in these studies, total energy intake was a confounder. In our study, associations were not observed regardless of energy-adjustment.

An FFQ is designed to rank individuals, not to measure absolute food consumption or nutrient intake, although the unit is g/day. In addition, fish consumption is usually over-reported in FFQ data.^26,27 Therefore FFQ-based intakes cannot be compared with recommendations. However, an advantage of an FFQ is that it evaluates the usual long-term diet, which is conceptually more important in epidemiological studies than absolute intake. In addition, the FFQ used in the present study has been calibrated against a 14-day food record and it was found to meet the requirements of epidemiological studies.²⁷

In the Fishermen Study, high serum DHA was associated with greater distress on the contrary to the hypothesis. However, this can be a coincidence due to the small number of participants or the very high levels of serum DHA in the Fisherman sub-study population. A rise in distress scores was seen after serum DHA concentration reached 6% both in men and women. Extremely high DHA concentrations may not be beneficial for mental well-being. Mental disorders such as depression have been associated with over-activity of the inflammatory response of the immune system increasing pro-inflammatory cytokines production.⁴² Omega-3 PUFAs have been suggested to inhibit some of these cytokines, especially of tumour necrosis alpha (TNF-α) and interleukin-1 beta (IL1-β).⁴³ However, in one recent study, DHA was reported to have pro-inflammatory effects by an induction of Th-1-like immune response.⁴⁴ For this reason, the study’s authors suggested that the treatment of depressed patients with fish oils should be avoided, since DHA could aggravate the immune regulatory system response in depression. Similarly, high fish consumption and omega-3 PUFA dietary intake were associated with more depressed mood in a Spanish study, which used a validated, semi-quantitative FFQ and energy adjustment.¹⁶ The authors suggest that one possible explanation might be the contamination of fish by mercury or other toxic compounds such as dioxins and polychlorinated biphenyls. More research is needed to assess whether a high intake of environmental contaminants could increase the risk of depressive symptoms among participants with high habitual fish consumption.

Instead of linear associations, we found some non-linear associations between fish consumption or omega-3 PUFA serum concentrations and psychological distress scores. These findings are in line with some previous studies suggesting that individuals with moderate consumption of fish or moderate omega-3 PUFA intake have a lower risk of mental disorder when compared with individuals with high consumption/intake.^13,16,45 In addition, stratified analyses revealed that some behavioural factors, such as physical activity, modified the studied associations. It is possible that fish consumption and omega-3 PUFA intake are associated with reduced psychological distress when lipid peroxidation has reduced serum levels of omega-3 PUFAs as a result of smoking or illness. However, the results concerning non-linear associations and effect modification were not consistent in any of the studied sub-groups or data sets. Therefore, the results of the present study suggest that there is no association between fish consumption or omega-3 PUFA intake and psychological distress. A failure to detect the hypothesized linear associations may simply reflect a more complex association between lifestyle and mood.

According to our results, there was no indication that high fish consumption or omega-3 PUFA intake would be associated with reduced distress either at the population level or in a high fish consumption group. Hence, the current evidence does not justify for recommendations to eat high amounts of fish to prevent symptoms of psychological distress.

Funding

Academy of Finland (project numbers 77008, 205324, 206950, 124286, 206950, 100499, 205585); Finnish Cancer Organisations; the Yrjö Jahnsson Foundation; the Juho Vainio Foundation; Academy of Finland Center of Excellence in Complex Disease Genetics (to J.K.) and The National Institutes of Health (NIAAA to J.K.), USA [grant AA-12502 to J.K. and Richard J Rose (P.I.)].

Acknowledgements

The authors would like to thank the volunteers and the research staff of the Fishermen Study, the Health 2000 Health Examination Survey and the Finntwin16 Study.

Conflict of interest: None declared.

KEY MESSAGES

• According to our results, high fish consumption does not prevent psychological distress.

• The associations of fish consumption and omega-3 PUFA intake with psychological distress could be non-linear.

• It is possible that the beneficial health effects of high fish consumption or omega-3 PUFA intake manifest only in individuals with detrimental health behaviour such as smoking or sedentary lifestyle.

• Very high serum concentrations of DHA may not be beneficial for mental well-being.