Essential Nutrients-Mood:  Related Paper

Essential Nutrients Menu

Essential Fatty Acids and Psychiatric Disorders


This book is out of print.  The following is a reprint of the above named chapter.
David F. Horrobin


There are strong reasons for considering the possible roles of essential fatty acids and prostaglandins in psychiatric disorders. Lipids account for 50-60% of the solid material in the brain and 15-20% of that lipid is in the form of essential fatty acids (EFAs) . Modifications of EFA intake or metabolism, especially if occurring at critical stages of development, may have profound effects on brain structure and function (1). The prostaglandins (PGs) derived from EFAs have been found to have dramatic effects on behaviour when injected into the cerebrospinal fluid (2). The PGs are also capable of modulating nerve conduction, neurotransmitter release, and post-synaptic transmitter actions (3-5). PGs of both 1 and 2 series are found in brain tissue and their levels change rapidly following changes of dietary EFA intake (6).

From a basic science point of view, it therefore seems reasonable that EFAs and PGs are worthy of investigation by those trying to understand the biochemical basis of mental illness. A number of clues have emerged from recent studies which indicate that such lines of study are likely to be productive. The purpose of this paper is to review briefly the status of PG and EFA studies in psychiatric disorders.


Abdulla et al (7) found that platelets from schizophrenics are severely defective in their ability to make PGE1. Since there may be a reciprocal relationship between levels of PGE1 and those of 2 series PGs (8), it is possible that in schizophrenia there is an excess of 2 series PGs coupled with a deficiency of 1 series PGs. A number of observations bear on this possibility:

1. Many of the clinical features of schizophrenia and its associated pathophysiology are consistent with the idea of a 1 series PG deficiency and a 2 series excess (3-5, 9).

2. Cerebrospinal fluid contains elevated levels of 2 series PGs in schizophrenics (10).

3. Red blood cell phosphatidyl-choline from schizophrenics contains elevated levels of arachidonic acid, the 2 series PG precursor, and reduced levels of dihomogamma-linolenic acid (DGLA), the 1 series PG precursor (11). Levels of linoleic acid were low in this study.

4. Alpha-linolenic acid levels are exceptionally high in schizophrenic blood (12).

We have made similar observations on EFA levels in schizophrenic blood. These suggest that in schizophrenia, the metabolism of the n6 series of EFAs from linoleic acid to arachidonic acid is exceedingly rapid, leading to an excess of arachidonic acid and deficiencies of linoleic acid and DGLA. On the other hand, metabolism of n3 series of EFAs seems exceptionally slow, leading to accumulation of alpha-linolenic acid (see the first review article in this volume). One possibility is that in schizophrenics the affinity of the desaturase enzymes for the n3 and n6 series of EFAs is reversed. Normally the enzymes have a higher affinity for the n3 series but in schizophrenics the n6 series would appear to be being metabolized faster. This would have profound effects on the brain structure since both n6 and n3 series of EFAs are found in large amounts, and also on brain function since the balance between the various series of PGs would be abnormal.

There have been three attempts to treat schizophrenia by raising 1 series PG formation. Chouinard et al (13) used penicillin which appears to activate PGE1 formation selectively. Vaddadi used penicillin in conjunction with Efamol (evening primrose oil containing gamma-linolenic acid) to act as a precursor for PGE1 (14). Parmigiani used Efamol in conjunction with captopril, a stimulator of PG biosynthesis (15). All three groups reported modest improvements which were of a type quite different from those seen with conventional neuroleptics. There was increased affect with more involvement in the life of the hospital, aspects of schizophrenia not influenced or made worse by the usual drugs. Vaddadi also noted some evidence of reduced tardive dyskinesia, a topic discussed further by him in this volume.


Some of Vaddadi's patients who had carried the label of `chronic schizophrenia' for many years in a long stay mental hospital became worse on being treated with Efamol. The worsening psychosis had some of the features of that associated with temporal lobe epilepsy which led Vaddadi to obtain electroencephalograms on these patients (16). Temporal lobe epilepsy was confirmed even though previous studies had failed to find it. On being switched to carbamazepine all three patients made dramatic recoveries and two were discharged from hospital. Obviously Efamol is contraindicated in temporal lobe epilepsy. However the effect may be of considerable clinical value as a diagnostic test allowing the identification of these patients. In a follow up study Vaddadi could find no effects of Efamol on the electroencephalogram of a group of normal individuals.


Abdulla et al (7) studied PGE1 formation in response to stimulation by ADP in platelets from manic and depressed individuals. At maximal ADP concentrations there were no differences between these two groups. However at half maximal stimulation, PGE1 levels in the depressed patients were significantly below normal, while those in the manic patients were significantly above normal. This suggests the possibility that manic-depression might involve disordered regulation of PGE1 biosynthesis.

This idea is reinforced by the observation that lithium at clinically relevant concentrations has a selective effect in inhibiting the mobilisation of dihomogamma-linolenic acid, the precursor of PGE1.  Lithium is the treatment of choice in manic depression. It has some acute anti-manic activity, no acute anti-depressive activity but the capacity to prevent both manic and depressive episodes. We have proposed that these effects can be explained by its actions on PGE1 (17). Suppose that mania is a condition in which the conversion of DGLA to PGE1 is uncontrolled. This conversion will continue until DGLA stores fall to a critical level at which time a lack of PGE1 will develop abruptly. If PGE1 is a determinant of mood, the mood swings of manic-depression could thus be explained. Lithium can suppress PGE1 formation and so would have an acute anti-manic effect. Lithium would have no acute anti-depressive effect because it would be acting in a situation in which PGE1 levels were already low. Lithium in the long term would prevent both mania and depression: by inhibiting excessive depletion of DGLA stores it would prevent the fall of PGE1 which would occur when those stores fell to a critical level.

Many patients being treated with lithium have great difficulty in tolerating the side effects. Lieb (19) reasoned that if lithium were administered to an EFA-depleted individual, then side effects might be expected because of the action of lithium on PGE1 formation. He found that administration of safflower oil or evening primrose oil greatly attenuated side effects of lithium and allowed patients previously intolerant of lithium to take it successfully.


Two quite separate actions of ethyl alcohol on the EFA/PG system have been described. First, alcohol is able to inhibit the delta-6-desaturase enzyme which converts linoleic acid to gamma-linolenic acid (GLA) (18, 20). Second, alcohol is able to modify the conversion of EFAs to PGs, greatly enhancing the formation of PGE1 from DGLA (21, 22) and modestly inhibiting the formation of thromboxane A2 from arachidonic acid (23, 24). These two actions suggest that alcohol could lead to a marked EFA deficiency. Its effects on PGE1 formation will tend to deplete precursor stores while its action on the desaturase will prevent replenishment of those stores from dietary linoleic acid. If this view is correct, the administration of GLA to by-pass the blocked step may be of great value in the treatment of alcoholism (24, 17). That this is indeed the case is suggested by the following observations:

1. The withdrawal syndrome in animals is markedly attenuated by GLA (see studies reported in detail in this volume by Rotrosen et al).

2. The withdrawal syndrome in humans may be modestly attenuated by the use of Efamol containing GLA (see report by Glen in this volume).

3. The fetal alcohol syndrome in rats may be prevented by administering evening primrose oil containing GLA (Efamol) simultaneously with alcohol during pregnancy (25).


The idea that hyperactivity in children may be associated with a defect in EFA metabolism has been proposed by the Hyperactive Children's Support Group in England (26). Four key observations suggested this idea:

1. The majority of hyperactive children come from families with a history of atopic disorders. Atopic disorders are known to respond to EFA supplementation (see paper by Wright et al in this volume).

2. Thirst is a symptom consistently reported by parents of hyper-active children. Thirst is an early feature of EFA deficiency in animals because of greatly increased water loss across the skin. EFAs are vital for maintaining the water impermeability of the skin.

3. Hyperactivity affects boys about three times more often than girls. Male animals are known to require about three times as much EFA as females to maintain growth. A partial EFA deficiency would therefore be expected to affect males much more than females.

4. A number of substances (tartrazine, salicylates and others) have been reported to have an adverse effect on behaviour in hyperactive children but not in normal children. Almost all the substances are known to be weak inhibitors of conversion of EFAs to PGs. They would not have much effect in the presence of normal amounts of EFA precursor but their influence could be critical if EFA stores were depleted.

The recent finding that atopic individuals have an apparent defect in the delta-6-desaturase enzyme which converts linoleic acid to gamma-linolenic acid (see paper by Horrobin and Manku, this volume) supports the idea of a defect in EFA metabolism in hyper-activity. A large number of children have now been tested with gamma-linolenic acid in the form of Efamol and about 80% seem to respond. An interesting feature of the 100 or so children studied so far is that the non-responders seem to be those with no history of atopic disorders in the family. Particularly important are the studies carried out by Walker (Norman Walker Associates).

He has been using a device to measure a child's tracking ability. He has developed objective, quantifiable measures of this and has shown that in hyperactive children the ability deteriorates much more dramatically on distraction than it does in normal individuals. He has now demonstrated that in such children Efamol produces a normalization of tracking ability within 1-2 weeks.


We are still at the very early stages of investigation of the relation-ships between EFAs, PGs and psychiatric illness. However the work done so far leaves little doubt that this hitherto largely ignored field will prove highly productive.


1. Crawford MA, Hassam AG, Stevens PA. Essential fatty acid requirements in pregnancy and lactation with special reference brain development. Progr Lipid Res 20: 31-40, 1982.
2. Horton EW, Main IHM. Differences in the effects of prostaglandin F2 alpha, a constituent of cerebral tissue, and prostaglandin El in conscious cats and chickens. Int J Neuropharmacol 4: 65-9, 1965.
3. Horrobin DF. Schizophrenia as a prostaglandin deficiency disease. Lancet 1: 936-7, 1977.
4. Horrobin DF. Schizophrenia: reconciliation of the dopamine, prostaglandin and opioid concepts and role of the pineal. Lancet 1: 529-31, 1979.
5. Horrobin DF. Prostaglandins and schizophrenia. Lancet 1: 706-7, 1980.
6. Hassam AG, Willis AL, Denton JP, et al. The effect of essential fatty acid deficient diet on the levels of prostaglandins and their fatty acid precursors in the rabbit brain. Lipids 14: 78-80, 1980.
7. Abdulla YK, Hamadah K. Effect of ADP on PGE formation in blood platelets from patients with depression, mania and schizophrenia. Br J Psychiatry 127: 591-5, 1975.
8. Horrobin DF. The regulation of prostaglandin biosynthesis: negative feedback mechanisms and the selective control of formation of 1 and 2 series prostaglandins: relevance to inflammation and immunity. Med Hypotheses 6: 687-709, 1980.
9. Horrobin DF, Ally AI, Karmali RA et al. Prostaglandins and schizophrenia: further discussion of the evidence. Psychol Med 8: 43-8, 1978.
10. Mathe AA, Sedvall G, Wiesel FA et al. Increased content of immunoreactive prostaglandin E in cerebrospinal fluid of patients with schizophrenia. Lancet 1: 16-18, 1980.
11. Hitzemann RJ, Garver DL. Abnormalities in membrane lipids associated with deficiencies in lithium counterflow. Society of Biological Psychiatry, New Orleans, June 1981.
12. Obi FO, Nwanze EAC. Fatty acid profiles in mental disease. Linolenate variations in schizophrenia. J Neurol Sci 43: 447-54, 1979.
13. Chouinard G, Horrobin DF, Annable L. Penicillin as an anti-
psychotic agent in schizophrenia. IRCS J Med Sci 6: 187, 1978.
14. Vaddadi KS. Penicillin and evening primrose oil in schizo-
phrenia. Prostaglandins Med 2: 77-80, 1979.
15. Parmigiani P. Evening primrose oil (Efamol) and captopril in
schizophrenia: a preliminary report (see this volume).
16. Vaddadi KS. Penicillin and evening primrose oil in schizo-
phrenia. Prostaglandins Med 2: 77-80, 1979.
17. Vaddadi KS. The use of gamma-linolenic acid and linoleic acid to differentiate between temporal lobe epilepsy and schizophrenia. Prostaglandins Med 6: 375-9, 1981.
18. Horrobin DF, Manku MS. Possible role of prostaglandin El in the affective disorders and in alcoholism. Br Med J 1: 1363-6, 1980.
19. Lieb J. Linoleic acid in the treatment of lithium toxicity and familial tremor. Prostaglandins Med 4: 275-9, 1980.
20. Reitz RC, Wang L, Schilling RJ. Effects of ethanol ingestion on the unsaturated fatty acids from various tissues. Progr Lipid Res 20: 209-13, 1982.
21. Nervi AM, Peluffo RO, Brenner RR. Effect of ethanol administration on fatty acid desaturation. Lipids 15: 263-8, 1980.
22. Rotrosen J, Mandio D, Segarnick D et al. Ethanol and prostaglandin El: biochemical and behavioural interactions. Life Sci 26: 1867-76, 1980.
23. Manku MS, Oka M, Horrobin DF. Differential regulation of the formation of prostaglandins and related substances from arachidonic acid and from dihomo-gamma-linolenic acid. 1. Effects of ethyl alcohol. Prostaglandins Med 3: 119-28, 1979.
24. Pennington SN, Smith CP. The effect of ethanol on thromboxane synthesis by blood platelets. Prostaglandins Med 2: 43-50, 1979.
25. Horrobin DF. A biochemical basis for alcoholism and alcohol-induced damage, including the fetal alcohol syndrome and cirrhosis: interference with essential fatty acid and prostaglandin metabolism. Med Hypotheses 6: 929-42, 1980.
26. Varma PK, Persaud TVN. Protection against ethanol-induced embryonic damage by linoleic and gamma-linolenic acid treatment. Prostaglandins Med 8: 641-5, 1982.
27. Colquhoun I, Bunday S. A lack of essential fatty acids as a possible cause of hyperactivity in children. Med Hypotheses 7: 673-9, 1981.