Joanne Bradbury1,2, Professor Stephen P. Myers 1, Chris Oliver 2,3
1 Australian Centre for Complementary Medicine, Education and Research (ACCMER), a joint venture between University of Queensland and Southern Cross University, NSW, Australia; 2 School of Natural and Complementary Medicine, Southern Cross University, NSW, Australia; 3 Blackmores Ltd, Australia
Abstract
Background: The purpose of this study was to confirm the evidence that low-fat diets are associated with stress and negative mood states.
Methods: Four self-report stress and mood measures were correlated with a crude dietary fat screen in 93 university staff that responded to an advertisement for a stress and dietary fats study. The screen was a modification of two previously validated dietary assessment questionnaires. The three stress measures were the Perceived Stress Scale (PSS), the Occupational Stress Inventory – Revised (OSI-R), and a visual analogue scale (VAS). The mood scale was the Positive and Negative Affect Scale (PANAS). Subjects completed the questionnaires at two measuring points, with a 10-week interval.
Results: At Time 1 there was an inverse correlation between fish intake and vocational strain (r= -0.30, p<0.01). That is, staff reporting high levels of fish intake also reported lower levels of work stress. At Time 2 there were inverse correlations between dietary fats and psychological strain (r= -0.27, p<0.05), physical strain (r= -0.26, p<0.05) and total personal strain (r= -0.28, p<0.05). That is, staff reporting higher dietary fat intakes also reported lower physical and mental stress.
Conclusion
This was a simple correlation study to ascertain whether dietary fat intakes are associated with stress. Several instruments were used. Only the most sophisticated instrument, the OSI-R strain subscales, had the sensitivity to detect significant relationships with the sample sizes used. While the specific relationships showed some variability over time, a significant association between dietary fats and stress was verified. Further studies are required to explore the reasons that the significant relationships showed variability over time, and whether dietary fats in general or specific fatty acids may be most ‘protective’ in stress.
Background
During the 1960s the rise in coronary heart disease was attributed to the increased consumption of saturated fatty acids. High cholesterol was identified as a risk factor and every attempt was made to reduce cholesterol. However, dietary induced cholesterol and fat reductions with the associated increase in carbohydrate intake have lately been associated with an increase in stress, violence and non-illness related mortality, such as suicide and fatal accidents [1, 2]. While these claims are controversial [3] and not conclusive [4], they are worthy of further investigation.
The major stress hormone released from the adrenal glands in response to stress is cortisol in humans and corticosterone in animals. In a randomised, placebo-controlled trial of 37 healthy men, a carbohydrate (glucose) load one hour before a psychosocial stressor significantly increased the cortisol response compared with fat and protein (p< 0.05) [5]. Carbohydrates seemingly intensified the stress response, where fat and protein had a relatively stabilising effect. These findings are consistent with animal studies demonstrating rapid elevations of corticosterone, insulin and glucose following carbohydrate rich meals [6]. Further, a fatty acid mixture given for 3 weeks prior to an acute stressor or an injection of corticosterone prevented the elevation in blood corticosterone in rats compared with controls (p<0.05) [7].
In a British study, 20 healthy volunteers were put on a high fat (41% fat, 46% carbohydrate, 14% protein) diet for one month. At the commencement of the second month, 10 were randomly selected to change to the trial diet, a low-fat/high carbohydrate diet (25% fat, 61% carbohydrate, 14% protein). After consuming a low-fat, high carbohydrate diet for one month, the group on the trial diet demonstrated significant (p=0.021) increases in anger-hostility, as measured by the Profile of Mood States (POMS) questionnaire. Additionally, tension-anxiety levels decreased in controls (p=0.025) as the study progressed, where the low-fat group showed no change [8].
Similar findings have been observed in animals. Thirty cynomolgus monkeys were fed either a ‘luxury diet’, where fat constituted 45% energy, or a ‘prudent diet’, where fat constituted 30% energy. After 22 months on the diets, 21 different behaviours were monitored. The only between-group difference observed was that monkeys on the prudent diet displayed significantly (p< 0.03) more contact aggression than monkeys on the luxury diet. Whether the animals on the prudent diet had exaggerated aggression or those on the luxury diet had decreased aggression was not clear. However, an association between low-fat diets and adverse mood effects was effectively demonstrated.
In an interesting study in 18 normal healthy adults in the UK, Zmarzty et al [9] demonstrated that a single high-fat/low-carbohydrate meal significantly changed the perception of pain. Subjects were asked to fast from midnight and attend the clinic at 9am for a meal of pancakes and lemon cordial. The meal was designed to be identical in calorie content and on measures of sensory stimulus such as taste. Subjects did not know that the macronutrient content of the meal would differ, and when later asked to rate the meals, no differences were detected. The energy content of the high fat meal was 54% fat, 40.8% carbohydrate, and 5.2% protein; and the high carbohydrate meal was 88% carbohydrate, 6.8% fat, and 5.2% protein. They were to attend on three separate mornings, where they were given one or another of the meals, and on one occasion no-meal was given.
At 1.5 hours after eating both meal conditions subjects reported feeling more drowsy and relaxed (p<0.05), than the no-meal condition. Although the authors report a trend for the difference to be higher in the high fat condition, there was no statistical significance in the difference between the two diet conditions. Most interesting, however, is that at 1.5 hours after the high-fat meal, subjects reported significantly lower levels of pain (p<0.01) when one hand was immersed in cold water than when they had no meal. While the carbohydrate meal was also protective against pain compared to the no-food condition, the analgesic properties of the high-fat/low-carbohydrate meal were significantly more potent (p=0.04) than that of the high-carbohydrate meal.
There are significant postprandial hormonal differences associated with isocaloric meals of varying macronutrient content. High-fat/low-carbohydrate meals (74% fat, 17% carbohydrates, 7% protein) are associated with significantly higher releases of cholecystokinin (CCK), than low-fat/high-carbohydrate meals (p=0.005); whereas high-carbohydrate/low-fat meals (11% fat, 81% carbohydrates, 7% protein) were associated with significantly higher release of insulin (p<0.0001), than high-fat/low-carbohydrate meals. In multiple regression analysis, CCK was positively correlated with fatigue (p=0.0062) and insulin was positively correlated with sleepiness (p=0.019) [10]. The mechanisms of action of CCK in sedation are unclear, but in the brain it may be involved in endogenous opioid pathways [11, 12].
Carbohydrate rich meals are associated with increases in the neurotransmitter serotonin. However, the ability of dietary carbohydrates to increase brain serotonin is dependant on the transport of its precursor, tryptophan, across the blood-brain barrier (BBB). This transport is often inhibited by competition with other large neutral amino acids (LNAA). In laboratory animals insulin facilitates the uptake of LNAA into skeletal muscles, leaving tryptophan free to cross into the brain [13]. In the brain it is converted into serotonin, which reduces pain and induces calm.
However, levels of dietary tryptophan are frequently not high enough to induce the levels of brain serotonin required to induce mood and behavioural changes in humans. In addition, as little as 4% dietary protein provides enough LNAAs for competitive inhibition at the BBB [9]. Experimental conditions in 334 Dutch university students demonstrated some benefit from an estimated 42% increase in dietary tryptophan. In order to achieve this, the carbohydrate rich diet was extremely low in protein (66.2% carbohydrate, 3.6 % protein, and 30.1% fat) compared with a protein rich diet (41% carbohydrate, 27% protein, 32% fats). Only a sub-group (those selected as highly proned to stress) demonstrated less cortisol reactivity (p=0.02) and lowered feelings of depression (p=0.025) in response to a stressful task after eating the carbohydrate-rich/protein poor diet [14].
There is a complex relationship between stress and serotonin (5-HT) in the brain. Basically stress increases cortisol, and elevated levels of cortisol results in reduced activity of serotonin, by desensitisation of 5-HT1A [15] and 5-HT1B [16] receptors. This implies that under stressful conditions there is not so much insufficient serotonin as an inability of these receptors to bind serotonin [17]. Dietary fat, but not carbohydrate, was shown to prevent desensitisation of 5-HT1A during stress[18].
In laboratory rats, a high fat diet (61% fat) administered for 2 months prior to exposure to a social stressor (being placed into the cage of a dominant rat), and a physiological stressor (being injected with the endotoxin, lipopolysaccharide (LPS)) was protective compared to a high carbohydrate diet (63% carbohydrates)[18]. After exposure to the social stressor, rats on the high fat diet had significantly less stress-induced raised body temperature (p=0.008), and were significantly less inhibited at night in their own cages as the rats fed on the high carbohydrate diet (p=0.01). When injected with LPS, the high fat group were less affected (p<0.05) and had reduced recovery times (p<0.01) than the high carbohydrate group. Thus dietary fats compared with carbohydrates were protective in both social and physiological stress in laboratory rats.
A possible mechanism for these effects may be the observed sparing of the serotonergic receptors from becoming desensitised to serotonin in the high fat group. Buwalda et al [18] demonstrated there was a significant between group difference (p=0.03) in the way the different groups responded to the 5-HT1A receptor agonist, 8-OHDPAT. Before the stressor, there were no differences in 8-OHDPAT-induced hypothermia between the groups. After the stressor, the high carbohydrate group were unable to respond, indicating that the 5-HT1A receptors were dysfunctional. There was no change in the ability of the high fat group to respond to the agonist after the stressor. Although this study was conducted in laboratory rats, the findings nonetheless support to the hypothesis that dietary fats may be protective in stress, mood and behaviour.
Finally, the nature and structure of the fatty acid have been implicated in mood and behaviour. In a study which considered epidemiological studies from many countries, higher levels of seafood consumption, indicative of consumption of omega-3 polyunsaturated fatty acids, were correlated with lower rates of depression [19]. Arguably the most essential structural role of omega-3 fatty acids is to maintain the integrity of cellular membranes, which control receptor functioning and ligand binding [20]. Omega 3 fatty acids have been implicated in serotonergic neurotransmission and have effects in humans comparable with pharmacological mood stabilisers [21].
Hibbeln and Salam [22] argue that dietary advise aimed at lowering cholesterol has led to a replacement of saturated fats with omega-6 fatty acids. A predominance of omega-6 fatty acids displaces the beneficial omega-3 fatty acids from cell membranes. Too much omega-6 coupled with insufficient omega-3 fatty acids has been related to depression, and a number of depression related disorders, including postpartum depression, alcoholism, and mood disorders.
There is now a wealth of literature linking deficiency of omega-3 fatty acids with adverse behavioural and mood effects. For instance, in a sample population of habitually violent and impulsive male offenders with antisocial personality, plasma phospholipid levels of the omega-3 fatty acid, docosahexaenoic acid (DHA), were significantly lower than controls (p<0.05), while the omega-6 fatty acid, arachidonic acid, metabolites prostaglandin E2 and thromboxane B2 levels were elevated (p<0.01) [23].
Further, in a sample of American boys aged 6-12 years, behaviour, learning and health problems were compared according to lower or higher essential fatty acid status in blood phospholipids. Those with lower levels of omega 3 fatty acids had higher scores on a range of behavioural, health and learning scales, including anxiety (p=0.008), hyperactivity (p=0.002), impulsivity (p =0.01), conduct (p =0.002), temper tantrums (p =0.002), sleeping difficulties (p =0.002), problems getting out of bed (p =0.0006), stomach-aches (p =0.03), and learning problems such as mathematics (p =0.04) and a lowered overall academic ability (p=0.03) [24].
Together, these findings suggest a possible protective role of dietary fats, particularly the omega-3 fatty acids, in stress and negative mood and behaviour states. The aim of the current study was to confirm an association between dietary fats and stress.
Methods
A longitudinal cross-sectional study correlating dietary fats intake with various measures of stress and mood was undertaken. Measurements were taken at baseline and after a 10-week interval. Subjects were required to complete the five self-report questionnaires at both measuring points. The study design is illustrated in Figure 1.
It should be noted that this study was part of a series of studies exploring the relationship between stress and fish oil. Thirty of the most stressed subjects were randomised into an intervention study, where fish oil or olive oil was supplemented for 6 weeks. While the intervention study will not be reported here, the fact that a sub group of subjects were treated differently should always be considered as a possible confounding factor.
Population
The study population consisted of 93 staff members from Southern Cross University.
An intra-staff email advertising the stress and dietary fat study invited all staff (no exclusions) employed by Southern Cross University, Australia. Recruitment was conducted over a 6-week period, which commenced in late April 2002. Questionnaires were re-administered 10 weeks later (Time 2) to all staff that had completed and returned the questionnaires at Time 1.
The Stress and Mood Measurement Instruments
1. Occupational Stress Inventory – Revised (OSI-R)
The OSI-R is a multi-dimensional questionnaire designed for the measurement of occupational stress. It has demonstrated reliability and validity as a measure of occupational stress [25]. The present study utilised the Personal Strain component, comprising of four subscales, each representative of different types of strain: (1) vocational strain (VS); (2) psychological strain (PSY); (3) interpersonal strain (IS); and (4) physical strain (PHS). Each subscale consists of 10 questions, and each question consists of a five-point range of possible responses from ‘rarely/never’ to ‘most of the time’. Total Personal Strain (PSQ) was also calculated as the average of the four strain subscales.
2. Perceived Stress Scale (PSS)
The PSS [26] is a global measure of perceived stress. The PSS is a simple tool, assessing the cognitive appraisals of life stress and coping resources. The PSS-10 has adequate reliability and validity as a stress measure [27]. It consists of 10 questions, each consisting of a five-point range of possible responses from ‘never’ to ‘very often’.
3. Visual Analogue Scale (VAS)
The VAS was a 10cm linear scale. The question asked: “How stressed has your life been over the past month?” The extreme left-hand side of the scale was marked ‘not stressed at all’ and the extreme right hand side was marked ‘extremely stressed’. The respondents’ mark on the scale is measured as distance from the left-hand side in centimetres.
4. Positive and Negative Affect Scales (PANAS)
The Positive and Negative Affect Scales (PANAS) consist of two 10-item scales, measuring two dimensions of mood. Each item is a word representing an emotion. Respondents were asked to estimate how frequently or intensely this emotion was experienced during the past month. Responses were according to a five-point range from ‘very slightly/not at all’ to ‘ very much’. The PANAS been demonstrated as a reliable and valid measure of mood [28 ].
A brief dietary screen for dietary fat intake
There are many different types of dietary fat. The common perception of fat is not likely to include fish or fish oil. In fact, most dietary fats assessment instruments do not differentiate between omega-6 and omega-3 polyunsaturated fatty acids. In order to incorporate these essential fatty acids two different dietary assessment measures were modified and combined to provide a simple dietary screen for (1) total fat intake, and (2) fish intake.
1. A simple score to measure dietary fat intake
Kinlay et al [29] developed a 12-item Short Questionnaire as a fast and inexpensive method of assessing group changes in dietary fat intake for intervention studies. The score estimates quantitative and qualitative dietary fat intake habits for an individual. The principle limitation of the Short Questionnaire is that it does not contain a complete list of foods containing fat and therefore is not sensitive for individual quantitative estimates. However, such a crude measure provides a simple guide to fat intake levels.
In a cross sectional study of 202 adults, the Short Questionnaire correlated well (saturated fat intake: r=0.6 p<0.05; total fat intake: r=0.49 p<0.05) with a nutrient intake standard measure (the CSIRO, Commonwealth Scientific Industrial Research Organisation, division of Human Nutrition Food Frequency Questionnaire)[29]. A number of minor modifications were made for the purposes of the current study: (1) the addition of full-fat yoghurt in with the ice-cream question, to more fully account for dairy intake; (2) the conversion of the question regarding chocolate intake from an open question to a closed question, to facilitate ease of scoring; (3) the inclusion of olive oil as an option for the fat in cooking question, and (4) the inclusion of olive/canola spread as an option for the question regarding usual spreads. A question was added to detail whether meat is regularly consumed in the diet. A second question was added to address the extent of convenience food consumption, such as take-away food or pre-packaged meals from the supermarket. The modifications were designed to differentiate between saturated, mono-unsaturated, trans and omega-6 poly-unsaturated fatty acids, in case the information was required for future analysis.
2. A simple score to measure dietary fish intake
Recently, Woods et al [30] validated a food frequency questionnaire measuring dietary fish intake against biological markers. In a study of 153 subjects, the correlations between total fish intake and serum DHA was 0.29 (95% CI 0.14-0.43). Steamed, grilled, baked and tinned fish (total non-fried fish) was correlated with DHA, 0.34 (95% CI 0.20 – 0.48). These findings indicate a clear relationship between self-report fish intake and serum DHA levels. The researchers concluded that these findings may be useful for ranking subjects according to likely serum levels when the resources are not available to permit the use of biological markers.
The food frequency questionnaire (FFQ) used by Woods et al [30] is comprehensive but includes three fish questions, which were used in the present study. These questions referred to amounts of (1) baked, grilled and steamed fish, (2) fried fish (including takeaways) and (3) tinned fish.
As scoring of the Woods FFQ involves computerised analysis for the whole FFQ, we developed a simple scoring method for the three questions, with the objective of crudely ranking subjects into categories based on frequency of fish consumption. In the modified scoring system, 0 was given for rarely, 1 was given for monthly, and 2 given for weekly. The modified scoring system has not been validated or tested for reliability.
Statistical Methods
An association would be statistically significant if the estimated value of r was greater than the critical value of r when α < 0.05. A significant finding would disprove the hypothesis that there is no relationship between dietary fats and stress. A significant correlation, therefore, would support the notion that there is an association between dietary fats and stress.
Results
The response rate to the recruitment advertisement was 13.6%. That is, of 810 full-time equivalent staff members, 110 responded to the email. Questionnaires were hand-delivered by the researcher, and returned through the university internal mail. At Time 1, ninety-three (93) staff members returned the completed questionnaires. At Time 2, sixty-eight (68) staff members returned the completed questionnaires.
The mean age in years of the respondents was 43.4 (SD 8.23). At Time 1, within the administration staff 44 were female and 10 were male; and the academic staff 24 were female and 15 were male. At time 2, within the administration staff 31 were female and 9 were male; and within the academic staff 16 were female and 12 were male.
Univariate analysis on all variables demonstrated normality and linearity. No serious violations or outliers were observed. Therefore the Pearson-moment correlational coefficient using the SPSS version 10.0 statistical software program was used for all correlational analysis.
Correlations between dietary fat intake, fish consumption and stress (PSS, VAS and OSI-R subscales)
There were two significant relationships between dietary fats and fish intake and the various stress and mood scales at Time 1 as shown in Table 1. The first was a negative correlation between fish consumption and vocational stress (-0.30, p<0.01). That is, people that reported higher levels of fish consumption also reported lower levels of vocational strain. The second was a negative relationship between fish consumption and total fat intake (-0.26, p<0.05). That is, lower levels of dietary fat intake were associated with eating more fish.
The relationships observed between dietary fats and stress at Time 1, however, were not repeated at Time 2, as shown in Table 2. That is, there was no relationship between fish consumption and vocational strain. However, there were negative correlations between dietary fat intake and three of the strain subscales: (1) physical strain (-0.26 p<0.05); (2) psychological strain (-0.26 p<0.05) and (3) total personal strain (-0.28 p<0.05). That is, people on low-fat diets experienced more strain, while those reporting higher levels of total fat intake reported less strain.
Discussion
Fish consumption and stress
There was a significant (p< 0.01) negative correlation between fish consumption and vocational strain (r=-0.30). This relationship implies that the people that eat more fish are less likely to be experiencing problems in quality or output of work, and have a more positive attitude towards work. It is also possible that fish consumers are a more health conscious group of people, and may exercise more and be more conscious of the health benefits of stress management. Nevertheless, the shared variance between fish consumption and vocational strain was 11%. That is, 11% of the variance in vocational strain was accounted for by fish consumption.
The relationship between fish consumption and vocational strain, however, was completely absent at Time 2. Several factors are worth consideration regarding interpretation of this discrepancy. Firstly, significance in the Pearson-moment correlational coefficient is sensitive to subject numbers. Secondly, Time 2 coincided with the beginning of the university semester. Subjects were returning from holidays and the vocational strains they reported mid-way through semester 1 might not have been present at the beginning of semester 2.
The latter interpretation for the discrepancy is interesting in light of the fact that the effects of fish oil have been shown to be effective in stressed but not non-stressed subjects [31]. Fish oil was found to prevent an increase in aggression in a study of 58 medical students during final examinations. The same research group conducted a similar study with the same design but timed to not coincide with a period of examination. The study consisted of 43 students and found no effects of fish oil on aggression. Perhaps the effects of fish were not evident at Time 2 because the subjects were not experiencing a level of stress sufficient to show an effect for the fatty acids.
A similar pattern has been observed with fish intake in mood. In depression, fish oil significantly (p=0.002) increased the duration of the remission periods [32], where dietary fish intake or fish oil did not effect mood in non-depressed individuals [4].
Fish consumption was also significantly (p< 0.01) negatively correlated with fat intake (r=- 0.26). This trend was also observed at Time 2, although significance was not reached. Significance of correlations is dependent on the sample size, which was smaller at Time 2. Low-fat diets were associated with increased fish consumption. This may be due to the use of fish as a protein replacement for meat or chicken. Generally, although not without exception, fish is accompanied by less saturated fat than other forms of meat. Nevertheless, fish consumption also accounts for a portion of the total dietary fat intake, as its oil is a rich source of polyunsaturated fatty acids.
Fat intake and stress
At Time 2, two of the personal strain variables of the OSI-R, psychological strain and physical strain were significantly negatively correlated with dietary fat intake. In addition, interpersonal strain, vocational strain and negative affect all had some correlation, although not significant with the current sample size. Total personal strain (the sum of the four personal strain subscales) was significantly negatively correlated with fat intake (-0.28, p< 0.05). Further, 8% of the variance in total personal strain was accounted for by dietary fat intake. That is, high levels of personal strain were associated with low-fat diets. Conversely, people with high fat intake report lower levels of personal strain.
Methodological issues
The results of the present study are inconclusive, as the same significant correlations were not observed at both measuring points. Possible confounds include: (1) the timing of the study; (2) the instrument used to measure dietary fat intake had not been validated or reliable; (3) the nutritional intervention study nested within this longitudinal correlational study meant that the dietary fats had been manipulated in approximately one third of the group between measurement points.
The timing of the questionnaires coincided with different types of strain. Time 1 (mid-semester 1) coincided with a lot more academic stress than Time 2 (beginning of Semester 2 and after the midyear break). That is, at Time 1 the main type of strain experienced was academic, but at Time 2, just after the break, the dominant expressions of stress were in the form of physical and mental strain. Perhaps there was a tendency of staff to exercise more, take outdoor holidays, spend more time with the family, or renovate the home. Any one or combination of these activities may lead to relatively increased physical and psychological demands. These confounds may have been avoided by making the second measuring point at mid-semester 2.
The OSI-R is a sophisticated instrument for measuring stress, strain and coping. Only one section was used in this study, the personal strain section. This section comprises of four different types of strain. While the academic strain was related to dietary fats at mid-semester, the physical and mental strains were associated with dietary fats at Time 2. Also at Time 2, total personal strain was related to dietary fats, a relationship probably due to the fact that total personal strain was highly correlated with psychological strain (0.9) at both measurement points, giving an exceedingly high level of shared variance.
The sensitivity of the OSI-R to detect different types of strain, combined with the study timing, could explain the different stains noted at the different measurement points. The main point is that stress was consistently significantly associated with lower dietary fat intakes.
Dietary fats and fish assessment
The fat intake assessment was a modified version of a validated instrument, where the modifications may have compromised the validity of the measure. Additionally, the original fat intake assessment had not been tested for reliability.
The measure was included in the study to give a crude estimation of fat intake to facilitate a simple correlational analysis between dietary fat intake and stress. The instruments used in the present study to estimate fat and fish intake were very crude. The questions were designed to give general estimates of fat and fish intake habits. They were not designed to give quantitative estimates. Further research with a validated and reliable dietary screening instrument is required to further substantiate these findings.
It should be noted that the FFQ proposed by Woods et al [10] demonstrated total fish consumption is a better indicator of serum omega-6 polyunsaturated fatty acids than DHA. This implies that total fish, as measured in the current study, is indicative of plasma omega-6 polyunsaturated fatty acids and is not a good indicator of plasma omega-3 levels.
The present study did not run separate analyses on the different types of fatty acids. For simplicity, all fats were included together for a single total fats intake. This methodology was used to crudely rank subjects according to usual fat intake habits. Similarly, all fish were included together for a total fish intake ranking. As fish is rich in polyunsaturated fatty acids, it was considered a type of dietary fat. The predominant type of fatty acid depends on the type of fish. Separate analysis of the particular types of fatty acids at this stage was beyond the specific aim of this study.
Sampling bias
Self-selection recruitment procedures invariably involves selection bias [33]. Those that chose to participate may be more motivated and positive. This may impact on appraisals of stress and mood because the sample may be biased in favour of perceiving stress from a more positive viewpoint. Additionally, the advertisement was aimed at university staff interested in stress research and dietary fat. This type of recruitment may have only attracted stressed staff, and therefore may not be representative of all staff members. Further, perhaps only those stressed staff interested in nutritional issues may have applied. On the other hand, staff may have excluded themselves due a heavy workload with little time to participate, or an unwillingness to disclose dietary fat habits.
Conclusion
There were several significant relationships found between dietary fats and stress. These findings depart significantly from that which would be expected under the null hypothesis, that there is no association between dietary fats and stress. Moreover, all associations found consistently demonstrate that higher stress and strain is associated with lower dietary fat intake.
Neither the simple visual analogue scale (VAS) nor the PSS were sensitive in detecting an association between stress and dietary fats and fish intakes. The OSI-R strain subscales, however, identified several significant relationships between dietary fats and fish intakes and various types of strain. Indeed, the OSI-R strain subscales demonstrated some interesting inverse relationships between dietary fat intake and psychological and physical strain.
Lower levels of psychological, physical and total personal strain were associated with higher dietary fat intakes (p<0.05). Moreover, low-fat diets were associated with higher levels of psychological, physical, and total personal strain. Another interesting inverse relationship was between fish intake and vocational strain (p<0.01). Lower levels of vocation strain were associated with higher levels of fish intake.
At both measurement points there were negative correlations between stress and dietary fats. The relationships varied only in the type of strain and the type of fat. It is important to note that a correlational analysis is not capable of indicating a cause-effect relationship. However, these findings are consistent with recent suggestions in the literature that dietary fats may play a protective or “adaptogenic” role in stress. The types of fatty acids and the types of stress and strain need further elucidation.
Acknowledgments
The authors would like to thank all the staff members from Southern Cross University that participated and generously gave their time to this research.
Competing interests
Chris Oliver is associated with Blackmores Ltd, a company that sells nutritional supplements including fish oil.
Longitudinal Cross-Sectional Study
Figure 1: Overview of the research
PSS=Perceived Stress Scale; VAS=Visual Analogue Scale;
OSI-R= Occupational Stress Inventory-Revised;
PANAS=Positive Affect and Negative Affect Scales
Table 1: Correlation coefficients for dietary fats, fish consumption, perceived stress, strain and mood at Time 1 (n=93).
Table 2: Correlation coefficients between dietary fat intake, fish consumption, perceived stress, strain and mood at Time 2 (n=68).
References
1. Muldoon MF, Manuck SB, Mattews KA: Lowering cholesterol concentration and mortality: A quantitative review of primary prevention trials. Br Med J 1990, 301:309-314.
2. Atrens DM: The questionable wisdom of a low-fat diet and cholesterol reduction. Soc Sci Med 1994, 39(3):433-447.
3. Rieger E: The diet-heart disease hypothesis: a response to Atrens. Soc Sci Med 1996, 42(9):1227-1233.
4. Wardle J, Rogers P, Judd P, Taylor MA, Rapoport L, Green M, Nicholson PK: Randomized trial of the effects of cholesterol-lowering dietary treatment on psychological function. Am J Med 2000, 108(7):547-553.
5. Gonzalez-Bono E, Rohleder N, Hellhammer DH, Salvador A, Kirschbaum C: Glucose but not protein or fat load amplifies the cortisol response to psychosocial stress. Horm Behav 2002, 41(3):328-333.
6. Wang J, Dourmashkin JT, Yun R, Leibowitz SF: Rapid changes in hypothalamic neuropeptide Y produced by carbohydrate-rich meals that enhance corticosterone and glucose levels. Brain Res 1999, 848(1-2):124-136.
7. Yehuda S, Rabinovitz S, Carasso RL, Mostofsky DI: Fatty acid mixture counters stress changes in cortisol, cholesterol, and impair learning. Int J Neurosci 2000, 101(1-4):73-87.
8. Wells AS, Read NW, Laugharne JD, Ahluwalia NS: Alterations in mood after changing to a low-fat diet. Br J Nutr 1998, 79(1):23-30.
9. Zmarzty SA, Wells AS, Read NW: The influence of food on pain perception in healthy human volunteers. Physiol Behav 1997, 62(1):185-191.
10. Wells AS, Read NW, Uvnas-Moberg K, Alster P: Influences of fat and carbohydrate on postprandial sleepiness, mood, and hormones. Physiol Behav 1997, 61(5):679-686.
11. Singh L, Oles RJ, Field MJ, Atwal P, Woodruff GN, Hunter JC: Effect of CCK receptor antagonists on the antinociceptive, reinforcing and gut motility properties of morphine. Br J Pharmacol 1996, 118(5):1317-1325.
12. Barbaz BS, Autry WL, Ambrose FG, Hall NR, Liebman JM: Antinociceptive profile of sulfated CCK-8. Comparison with CCK-4, unsulfated CCK-8 and other neuropeptides. Neuropharmacology 1986, 25(8):823-829.
13. Rouch C, Nicolaidis S, Orosco M: Determination, Using Microdialysis, of Hypothalamic Serotonin Variations in Response to Different Macronutrients. Physiol Behav 1998, 65(4-5):653-657.
14. Markus R, Panhuysen G, Tuiten A, Koppeschaar H: Effects of food on cortisol and mood in vulnerable subjects under controllable and uncontrollable stress. Physiol Behav 2000, 70(Aug Sep, (3 4)):333-342.
15. Laaris N, Le Poul E, Hamon M, Lanfumey L: Stress-induced alterations of somatodendritic 5-HT1A autoreceptor sensitivity in the rat dorsal raphe nucleus–in vitro electrophysiological evidence. Fundam Clin Pharmacol 1997, 11(3):206-214.
16. Bolanos JF, Manhaes de Castro RM, Seguin L, Cloez TI, Monneret V, Drieu K, Fillion G: Effects of stress on the functional properties of pre- and postsynaptic 5-HT1B receptors in the rat brain. Eur J Pharmacol 1995, 294(2-3):531-540.
17. Czyrak A, Mackowiak M, Chocyk A, Fijal K, Gadek Michalska A, Wedzony K: 8-OHDPAT-induced disruption of prepulse inhibition in rats is attenuated by prolonged corticosterone treatment. Neuropsychopharmacology 2003, 28(7):1300-1310.
18. Buwalda B, Blom WA, Koolhaas JM, van Dijk G: Behavioral and physiological responses to stress are affected by high-fat feeding in male rats. Physiol Behav 2001, 73(3):371-377.
19. Noaghiul S, Hibbeln JR: Cross-national comparisons of seafood consumption and rates of bipolar disorders. Am J Psychiatry 2003, 160(12):2222-2227.
20. Lachman HM, Papolos DF: Abnormal signal transduction: a hypothetical model for bopilar affective disorder. Life Sci 1989, 45:1413-1426.
21. Stoll AL, Locke CA, Marangell LB, Severus WE: Omega-3 fatty acids and bipolar disorder: a review. Prostaglandins Leukot Essent Fatty Acids 1999, 60(5-6):329-337.
22. Hibbeln J, Salem N, Jr: Dietary polyunsaturated fatty acids and depression: When choleterol does not satisfy. Am J Clin Nutr 1995, 62(1):1-15.
23. Virkkunen M, F. , Jenkins M, E. , Horrobin D, F., Jenkins D, D. , Manku M, S.: Plasma phospholipid essential fatty acids and prostaglandins in alcoholic, habitually violent and impulsive offenders. Biol Psych 1987, 22:1087-1096.
24. Stevens LJ, Zentall SS, Abate ML, Thomas K, Burgess JR: Omega-3 fatty acids in boys with behaviour, learning, and health problems. Physiol Behav 1996, 59(4/5):915-920.
25. Osipow SH: Occupational Stress Inventory Revised Edition (OSI-R) – Professional Manual. In: PAR – Psychological Assessment Resources, Inc. Florida; 1981.
26. Cohen S, Kamarck T, Mermelstein R: A global measure of perceived stress. J Health Soc Beh 1983, 24(4):385-396.
27. Cohen S: Measures of psychological stress. In. http:www.macses.ucsf.edu/Research/Pscyosocial/notebook/stress.html; 2000.
28. Watson D, Clark L, Tellegen A: Development and validation of brief measures of positive and negative affect: The PANAS scales. J Pers Soc Psychol 1988, 54(6):1063-1070.
29. Kinlay S, Heller R, Halliday J: A simple score and questionnaire to measure group changes in dietary fat intake. Prev Med 1991, 20:378-388.
30. Woods R, Stoney R, Ireland P, Baily M, Raven J, Thien C, Walters E, Abramson M: A valid food frequency questionnaire for measuring dietary fish intake. Asia Pacific J Clin Nutr 2002, 11(1):65-61.
31. Hamazaki T, Sawazaki S, Nagasawa T, Nagao Y, Kanagawa Y, Yazawa K: Administration of docosahexaenoic acid influences behavior and plasma catecholamine levels at times of psychological stress. Lipids 1999, 34 Suppl:S33-S37.
32. Stoll AL, Severus WE, Freeman MP, Rueter S, Zboyan HA, Diamond E, Cress KK, Marangell LB: Omega 3 fatty acids in bipolar disorder: a preliminary double-blind, placebo-controlled trial. Arch Gen Psychiatry 1999, 56(5):407-412.
33. Contrada R, Krantz D: Measurement bias in health psychology research designs. In: Stress and Health: Issues in Research Methodology. Edited by Kasl SV, Cooper CL. Chichester: John Wiley and Sons; 1987: 57-78.
Editorial Commentary
In the current issue of the International Journal of Naturopathic Medicine, Bradbury et al describe an inverse correlation between fish intake and work related stress. In addition, these investigators found that employees working in an educational institution who report higher dietary fat intakes have decreased levels of self-reported physical and mental stress. The findings of Bradbury et al make a significant contribution to a rapidly growing body of international research which connects nutritional intake and behaviour.
Recent epidemiological, experimental and clinical studies suggest that the influence of nutritional factors on mental health is currently under-appreciated[1]. A recent study involving over 3500 young adults living in North American cities found that high dietary intake of fish rich in omega-3 fatty acids, and docosahexaenoic acid (DHA) in particular, was associated with decreased hostility[2]. In a 9-month study published in the British Journal of Psychiatry, researchers investigated the effects of a basic multivitamin/mineral formula and fish oil (with gamma-linolenic acid) in a prison for young adults. Supplementation with the essential fatty acids and vitamins/minerals led to a 26.3% reduction in antisocial and violent behaviour vs. those taking placebos[3]. Another recently published study showed that among cocaine addicts, lower levels of total omega-3 fatty acids and DHA are found in those with a past history of aggression vs. cocaine addicts without aggressive history[4]. Interestingly, low plasma cholesterol predicts relapse in detoxified cocaine addicts[5].
The current research by Bradbury et al has implications when considering pharmacotherapy. The emerging research on dietary fat and behaviour suggests that the widespread use of statin drugs to lower cholesterol should be examined more closely. Japanese researchers found in a large study (n=13,571) those with the lowest cholesterol levels were most likely to be in a depressive state[6]. In addition, patients who attempt suicide have been found to have significantly lower cholesterol levels than psychiatric and healthy controls[7]. Low cholesterol levels are associated with low serotonergic neurotransmission, which may lead to depression[8]. Furthermore, selective serotonin re-uptake inhibitors, the first-line pharmacotherapy in depression, have been shown experimentally to lead to decreased fat intake[9,10]. This may be one factor in treatment resistance depression, at the very least, it indicates that those on prescription antidepressants require essential fatty acid supplementation.
The current findings by Bradbury et al do not prove causation, and obviously social, economic and other confounding factors may interfere with interpretation. Still, the results are in line with previous reports, and should encourage further investigation in this exciting area of research.
Alan C. Logan, ND, FRSH
Nutrition Editor
References:
- Horrobin DF. Food, micronutrients, and psychiatry. Int Psychogeriatr 2002 Dec;14(4):331-4.
- Iribarren C, Markovitz JH, Jacobs DR Jr, Schreiner PJ, Daviglus M, Hibbeln JR. Dietary intake of n-3, n-6 fatty acids and fish: relationship with hostility in young adults–the CARDIA study. Eur J Clin Nutr 2004 Jan;58(1):24-31.
- Gesch CB, Hammond SM, Hampson SE, Eves A, Crowder MJ. Influence of supplementary vitamins, minerals and essential fatty acids on the antisocial behaviour of young adult prisoners. Randomised, placebo-controlled trial. Br J Psychiatry 2002 Jul;181:22-8.
- Buydens-Branchey L, Branchey M, McMakin DL, Hibbeln JR. Polyunsaturated fatty acid status and aggression in cocaine addicts. Drug Alcohol Depend 2003 Sep 10;71(3):319-23.
- Buydens-Branchey L, Branchey M. Association between low plasma levels of cholesterol and relapse in cocaine addicts. Psychosom Med 2003 Jan-Feb;65(1):86-91.
- Terao T, Iwata N, Kanazawa K, Takano T, Takahashi N, Hayashi T, Sugawara Y. Low serum cholesterol levels and depressive state in human dock visitors. Acta Psychiatr Scand 2000 Mar;101(3):231-4.
- Kim YK, Lee HJ, Kim JY, Yoon DK, Choi SH, Lee MS. Low serum cholesterol is correlated to suicidality in a Korean sample. Acta Psychiatr Scand 2002 Feb;105(2):141-8.
- Terao T, Nakamura J, Yoshimura R, Ohmori O, Takahashi N, Kojima H, Soeda S, Shinkai T, Nakano H, Okuno T. Relationship between serum cholesterol levels and meta-chlorophenylpiperazine-induced cortisol responses in healthy men and women. Psychiatry Res 2000 Oct 30;96(2):167-73.
- Heisler LK, Kanarek RB, Gerstein A. Fluoxetine decreases fat and protein intakes but not carbohydrate intake in male rats. Pharmacol Biochem Behav 1997 Nov;58(3):767-73.
- Heisler LK, Kanarek RB, Homoleski B. Reduction of fat and protein intakes but not carbohydrate intake following acute and chronic fluoxetine in female rats. Pharmacol Biochem Behav 1999 Jul;63(3):377-85.