Schizophrenia c
Research,
1992 Elsevier
SCHIZO
6 (1992)
1-l
Science Publishers
B.V. All rights reserved
0920-9964/92/$05.00
00194
Fatty acids in plasma phospholipids and cholesterol esters from identical twins concordant and discordant for schizophrenia C. Bates’, ’ Thurston-Mason
Community
Mental
(Received
Health
28 January
D.F.
Horrobin2
and
K. Ells2
Center, Olympia, WA 98507, U.S.A.. and 21Sfamol Research Kentville. Nova Scotia B4N 4HB. Canada
1991, revised received
25 April
1991, accepted
30 April
Institure,
P.O. Box 818,
1991)
The fatty acid compositions of plasma phospholipids and cholesterol esters were measured in 18 pairs of twins discordant for schizophrenia and 20 pairs concordant for schizophrenia. In the twins discordant for schizophrenia the only significant abnormalities were elevations of adrenic (22 : 4n-6) and docosapentaenoic (22 : 5n6) acids in the schizophrenic twins. These fatty acids have also recently been reported to be elevated in brains from schizophrenics. The twins concordant for schizophrenia showed many differences from the normal discordant twins. 22: 4n-6 and 22: 5n6 were even more abnormal than in the schizophrenic discordant twins. In addition, linoleic acid was significantly reduced, an abnormality which has been found consistently in other schizophrenic populations. These observations are consistent with the concept that unsaturated fat metabolism may be abnormal in schizophrenia. Key words:
Twins; Essential
fatty acids; Adrenic
acid; Linoleic
INTRODUCTION
The majority of cerebral tissue is made up of lipids. The balance between saturated, monounsaturated and essential polyunsaturated fatty acids in membranes determines membrane microviscosity (Biagi et al., 1991) and the behaviour of membranebound proteins such as receptors, ion channels and ATPases (Ho and Cox, 1982; Clerc-Hoffmann et al., 1983; Goodfriend and Ball, 1986; Mitsuhashi et al., 1986; Horrobin and Manku, 1989). Certain derivatives of the essential fatty acids such as prostaglandins (PGs) and leukotrienes (LTs) have profound effects on many neuronal functions such as nerve conduction, transmitter release and receptor function (Horton, 1964; Hedquist, 1976; Horrobin et al., 1977, 1978; Manku et al., 1977; Horrobin, 1988). There is a physiological antagonCorrespondence IO: D.F. Horrobin, Efamol Research Institute, P.O. Box 818, Kentville, Nova Scotia B4N 4HB, Canada (Tel.: 902-678-5534).
acid; (Schizophrenia)
a derivative of dihomo-yism between PGEi, linolenic acid (DGLA, 20: 3n6) and dopamine, an excess of one being able to counteract an excess of the other (Masek et al., 1976; Blosser, 1978). The main saturated fatty acids in membrane phospholipids are palmitic (16 : 0) and stearic (18 : 0), and the main monounsaturated fatty acid is oleic (18 : I n9). The polyunsaturated essential fatty acids are the main determinants of membrane fluidity, having effects proportional both to their concentration and to the number of double bonds in the molecule. The essential fatty acids are derived from linoleic acid (18 : 2n-6) and cc-linolenic acid (18 : 3n-3) by the pathways shown in Fig. I. In the shorthand notation, the number before the column indicates the number of carbon atoms in the molecule, the number after the column indicates the number of double bonds and the n number indicates the family (n-6 for linoleic and n-3 for x-linolenic acids). Several investigators have reviewed the evidence for the idea that there may be abnormalities in
n-6 Linoleic (LA)
Gamma-linolenic
EFAs
n-3
16:2n-6
EFAs
16:3n-3
Alpha-linolenic (ALA)
delta-6-desaturase 1 16:3n-6
1 18:4rl-3
1 20:3n-6
1 20:4n-3
(GLA) Dihomogammalinolenic (DGLA)
delta-Sdesaturase Arachidonic
1 20:4n-6
1 20:5n-3
1 22:4n-6
& 22:5n-3
(AA) Adrenic
Eicosapentaenoic (EPA)
&&a-4-desaturasel 22:5n-6
22:6n-3
Docosahexaenoic
Fig. I
fatty acid and eicosanoid metabolism in various psychiatric disorders, notably schizophrenia (Feldberg, 1976; Horrobin, 1977, 1979, 1985, 1989; Horrobin et al., 1978; Van Kammen et al., 1989). It has recently been demonstrated that in the phospholipid (P), P-ethanolamine, from the frontal cortex of schizophrenics there are reduced concentrations of linoleic acid and its IS-carbon and 20carbon metabolites, balanced by a corresponding increase of the 22-carbon fatty acids (Horrobin, 1989; Horrobin et al., 1991). In plasma phospholipids, linoleic acid levels are consistently lower in schizophrenics from the same geographical area than in controls (Horrobin et al., 1989; Kaiya et al., 1991). This has been shown in Scottish. Irish, English and Japanese populations. Other fatty acid differences between normals and schizophrenics have been inconsistent. DGLA levels were significantly elevated in Japanese schizophrenics, nonsignificantly elevated in Irish schizophrenics and significantly below normal in Scottish and English schizophrenics. DGLA was found to be significantly elevated in violent criminals (Virkunnen et al., 1987). Identical twins offer unique opportunities for the study of biological markers for schizophrenia. When such twins are concordant for schizophrenia, the genome must make phenotypic schizophrenia particularly likely. When they are discordant, the effect of the genome has been over-ruled by environmental factors in one twin: whether the environment has made the normal genome schizophrenic or the schizophrenic genome normal is, of course, a matter for much debate. As part of an ongoing study of schizophrenia conducted at the National Institute of Mental Health, Washington, plasma
samples were collected from 28 pairs of identical twins. Ten of those pairs were concordant for schizophrenia, with both twins exhibiting clinical features of the disease. 18 pairs were discordant, with one exhibiting schizophrenia and the other not. The first report from this study has already appeared (Suddath et al., 1990). It demonstrated that in the pairs discordant for schizophrenia, the schizophrenic twins consistently had slightly enlarged cerebral ventricles and abnormalities in the hippocampal region. We now report on the fatty acid patterns in the plasma phospholipids and cholesterol esters in these patients.
PATIENTS
AND
METHODS
The design of the twin study, the diagnostic criteria, and information about the patients have been described in detail elsewhere (Suddath et al., 1990). In brief the diagnosis was based on revised DSM-III criteria, and in all the discordant pairs the minimum duration of discordance was 4 years. At the time of sampling, most of the schizophrenics were also being treated with neuroleptics. The characteristics of the concordant and discordant pairs are shown in Table 1. Of the ten concordant pairs, five were female as were two of the 18 discordant pairs. We thank Dr. E. Fuller Torrey of the National Institute of Mental Health, St. Elizabeth’s Hospital, Washington. DC 20032, for providing samples. Plasma samples were frozen, coded and sent by air to the Efamol Research Institute. The biochemists at the Institute were entirely unaware as to which samples came from which group. On completion of the analyses the results were sent to Dr. Bates and only then was the code broken. The samples were analysed for their fatty acid composition as described previously (Manku et al., 1983). In brief, lipids were extracted and separated into the various fractions by thin layer chromatography. The phospholipid and cholesterol ester bands were scraped off, transmethylated and the fatty acid composition of the methyl ester derivatives determined by gas chromatography. Each fatty acid in each patient group was compared statistically with the levels of that fatty acid in the normal twins. This comparison was usually
3
by Student’s t test but when fatty acid levels were below 1 mg/lOO mg and not symmetrically distributed. Mann-Whitney non-parametric tests were used.
RESULTS
The results of the phospholipid fatty acid analyses are shown in Table 2. The main features to which we wish to draw attention are the following: (1) linoleic acid levels in the discordant pairs were similar in both normals and schizophrenics but significantly below the discordant normals in the concordant schizophrenics; (2) y-linolenic and dihomo-y-linolenic acid levels were also similar in both normals and schizophrenics in the discordant pairs, but elevated in the concordant schizophrenics; (3) the most abnormal fatty acids were the 22 carbon n-6 compounds which were significantly elevated in both schizophrenic groups. However, the discordant schizophrenic twins were about half way between the concordant twins and the normals; (4) the n-3 fatty acids were normal in the discordant schizophrenics except for docosahexaenoic acid which was marginally low. Both r-linolenic and eicosapentaenoic acids were slightly elevated in the concordant schizophrenics with docosahexaenoic acid again marginally low; (5) other fatty acids were normal or near normal with a slight tendency for stearic acid to be elevated in both schizophrenic groups. The results in the cholesterol ester fraction are shown in Table 3. There were few significant abnormalities although there were some trends and differences of borderline significance (2~ < 0.1) almost entirely in the concordant schizophrenic group. In this group linoleic acid was again low while y-linolenic and dihomo-y-linolenic acids were slightly high. Oleic acid was elevated in the concordant schizophrenics.
DISCUSSION
Overall, the results support the view that there are abnormalities of fatty acid metabolism in schizo-
phrenics. This may prove to be important in view of the important roles played by those fatty acids in modulating neuronal function and in neuronal structure. However, as in all similar studies, one must insert the caveat that we cannot be certain whether some or all changes noted might be attributable either to prolonged drug treatment or to changes in lifestyle occurring because of the illness. The changes might also be attributable to some form of biochemical compensatory response to the illness. Ultimately these issues will be addressed only when the effect on the illness is investigated by manoeuvres which restore the fatty acid abnormalities to or towards normal. If such normalisation improves the clinical condition than the fatty acids are likely to be playing some causal role: if such intervention has no effect then the fatty acid changes are likely to be secondary. There were only a few differences between the discordant schizophrenics and their normal twins and these differences were largely confined to the 22-carbon fatty acids which play important roles in brain structure. The two fatty acids which were found to be elevated in the plasma phospholipids are the two whose levels were raised in frontal cortex p-ethanolamine (Horrobin, 1989; Horrobin et al., 1991). This raises the possibility of an abnormality in the conversion of precursors to the 22 carbon derivatives and in their transfer from plasma to cerebral tissue. Enzymes which might be involved are those producing elongation and desaturation of fatty acids and those incorporating fatty acids into membrane phospholipids and other complex lipids. Because they are relatively minor components of the plasma phospholipids, little attention was paid to adrenic and docosapentaenoic acids in previous studies (Horrobin et al., 1989; Kaiya et al., 1991). However, the findings in the present study are consistent with the trends in the previous investigations. Most of the fatty acid levels in the concordant schizophrenics were different from those in both the discordant schizophrenics and the normals. The same 22-carbon n-6 fatty acids were elevated as in the discordant schizophrenics, but the degree of abnormality was considerably greater. Linoleic acid was significantly reduced, consistent with all the reports to date on plasma from Japanese,
4 TABLE
1
Characteristics
of the patients
whose blood samples were investigated
in this study
No.
Age
Sex
Race
111.01 112.01 113.01 115.01 117.01 118.01
44 28 30 35 36 26
M M F F M F
Cauc. Cauc. Cauc. Cauc. Cauc. Cauc.
119.02
27
F
Cauc.
9
120.01
24
M
Cauc
2
122.01 125.02
34 28
M F
Cauc. Cauc.
20 9
301.01
31
M
African
Amer.
n/a”
301.02
31
M
African
Amer.
n/a
302.01 303.01
38 24
M M
Cauc. Cauc.
nja
303.02
24
M
Cauc.
n/a
305.01
25
M
Cauc.
da
305.02
25
M
Cauc.
n/a
308.01
41
F
Cauc.
n/a
308.02
41
F
Cauc
n/a
309.01
31
M
Cauc.
Years discordant
n/a
Medications
Mellaril 100 mg/day Prolixin D 37.5 mg/q 2 weeks Off all medication for 7 weeks Mellaril 250 mg/day Haldol 80 mg/day Lithium 1200 mg/day Tegretol 1200 mg/day Haldol 60 mg/day Tofranil 100 mg/day Navane 10 mg/day Desipramine 100 mg/day Off all medications for k 3 year Haldol 25 mg/day Haldol 50 mg. i.m. q 2 weeks Cogentin 2 mg/day Ativan PRN Prolixin (unk. dosage) Lithium (unk. dosage) Ascendin 200 mg/ IO mg Cogentin 600 mg/p.r.n. Navane IO-20 mg/day Ascendin (unk. dosage) Cogentin (unk. dosage) Prolixin 2.5 mg/day Loxitane 75 mg/day Artane 10 mg/day Lithium 1200 mg/day Azulfidine (unk. dosage) Cogentin 2 mg/day Stelazine 20 mg/day Sulfadine 4000 mgiday Tegretol 1200 mg/day Prednisone 40 mg/day Trilafon 12 mg/day Rivitral 0.5 mg/b.i.d. Trilafon 36 mg/day Kendrin 5 mg/t.i.d. Nazinan 37.5 mg/day Haldol 50 mg/i.m. q 4 weeks Haldol 80 mg/day Lithium 1200 mg/day Artane (unk. dosage) Loxitane 100 mg/day Artane (unk. dosage) Haldol 15 mg/day Benadryl (unk. dosage) Lithium 1200 mg Surmentil 200 mg/h.s. Navane 400 mg/day Mellaril 25 mg/q.i.d. p.r.n. L-tryptophan 500 mg/day Amatadine 100 mg/q.i.d.
TABLE
1 (continued)
sex
No.
YlTUS
RUCt?
Medications
discordant Cauc.
n/a
27
Cauc.
n/a
312.02
21
Cauc.
nia
313.01 313.02 314.01 314.02
26 26 29 29
Cauc. Cauc. Cauc. Cauc.
n/a n/a n/a
316.01
22
M
Cauc.
n/a
316.02
22
M
Cauc.
n/a
309.02
31
312.01
“Non-applicable TABLE
M
n/a
Stelazine 20 mg/day Lithium 900 k 122 mg on alt. days Klonopin 0.25 mg/t.i.d. Artane 5 mg/day Stelazine 20 mgiday Cogentin 2 mgiday Stelazine 20 mg/day Cogentin 2 mg!day Thorazine 400 mg/day Unknown Unknown Trilafon 16 mg/day Lithium 1500 mg/‘day Benadryl 50 mg/t.i.d. Stelazine 25 mg/day Artane 4 mg/day Mellarin 150 mg/day
since those twins were concordant
2
Concentrations of fatty acids (mg/lOO schizophrenic twins of the Jirst group. together Results are expressed as means&SD. concentrations of the fatty acids were Whitney ranking test.
mg total lipid) in the plasma phospholipid fraction of the normal monozygotic iwins. oJ the qf‘ the schizophrenic twins who were both schkophrenic and of all ~hr schtophrenics grouped
Fatty acid
Normal,
18 : 2n-6, linoleic 18 : 3n-6, y-linolenic 20 : 3n-6, dihomo-y-linolenic 20 : 4n-6, arachidonic 22 : 4n-6, adrenic 22 : 5n-6, docosapentaenoic 18 : 3n-3, cc-linolenic 20 : 5n-3, eicosapentaenoic 22 : 5n-3, docosapentaenoic 22 : 6n-3, docosahexaenoic 16 : 0, palmitic 18 : 0, stearic 18: I, oleic
30.4k4.5 0.07+0.15 2.8kO.8 12.7k2.7 0.14*0.17 0.17*0.18 0.18kO.20 0.95&0.51 1.1+1.7 4.0 + 2.6 25.6k2.2 8.1kl.3 I l.Ok2.3
Sta/istical ‘2p
Research,
1992 Elsevier
SCHIZO
6 (1992)
1-l
Science Publishers
B.V. All rights reserved
0920-9964/92/$05.00
00194
Fatty acids in plasma phospholipids and cholesterol esters from identical twins concordant and discordant for schizophrenia C. Bates’, ’ Thurston-Mason
Community
Mental
(Received
Health
28 January
D.F.
Horrobin2
and
K. Ells2
Center, Olympia, WA 98507, U.S.A.. and 21Sfamol Research Kentville. Nova Scotia B4N 4HB. Canada
1991, revised received
25 April
1991, accepted
30 April
Institure,
P.O. Box 818,
1991)
The fatty acid compositions of plasma phospholipids and cholesterol esters were measured in 18 pairs of twins discordant for schizophrenia and 20 pairs concordant for schizophrenia. In the twins discordant for schizophrenia the only significant abnormalities were elevations of adrenic (22 : 4n-6) and docosapentaenoic (22 : 5n6) acids in the schizophrenic twins. These fatty acids have also recently been reported to be elevated in brains from schizophrenics. The twins concordant for schizophrenia showed many differences from the normal discordant twins. 22: 4n-6 and 22: 5n6 were even more abnormal than in the schizophrenic discordant twins. In addition, linoleic acid was significantly reduced, an abnormality which has been found consistently in other schizophrenic populations. These observations are consistent with the concept that unsaturated fat metabolism may be abnormal in schizophrenia. Key words:
Twins; Essential
fatty acids; Adrenic
acid; Linoleic
INTRODUCTION
The majority of cerebral tissue is made up of lipids. The balance between saturated, monounsaturated and essential polyunsaturated fatty acids in membranes determines membrane microviscosity (Biagi et al., 1991) and the behaviour of membranebound proteins such as receptors, ion channels and ATPases (Ho and Cox, 1982; Clerc-Hoffmann et al., 1983; Goodfriend and Ball, 1986; Mitsuhashi et al., 1986; Horrobin and Manku, 1989). Certain derivatives of the essential fatty acids such as prostaglandins (PGs) and leukotrienes (LTs) have profound effects on many neuronal functions such as nerve conduction, transmitter release and receptor function (Horton, 1964; Hedquist, 1976; Horrobin et al., 1977, 1978; Manku et al., 1977; Horrobin, 1988). There is a physiological antagonCorrespondence IO: D.F. Horrobin, Efamol Research Institute, P.O. Box 818, Kentville, Nova Scotia B4N 4HB, Canada (Tel.: 902-678-5534).
acid; (Schizophrenia)
a derivative of dihomo-yism between PGEi, linolenic acid (DGLA, 20: 3n6) and dopamine, an excess of one being able to counteract an excess of the other (Masek et al., 1976; Blosser, 1978). The main saturated fatty acids in membrane phospholipids are palmitic (16 : 0) and stearic (18 : 0), and the main monounsaturated fatty acid is oleic (18 : I n9). The polyunsaturated essential fatty acids are the main determinants of membrane fluidity, having effects proportional both to their concentration and to the number of double bonds in the molecule. The essential fatty acids are derived from linoleic acid (18 : 2n-6) and cc-linolenic acid (18 : 3n-3) by the pathways shown in Fig. I. In the shorthand notation, the number before the column indicates the number of carbon atoms in the molecule, the number after the column indicates the number of double bonds and the n number indicates the family (n-6 for linoleic and n-3 for x-linolenic acids). Several investigators have reviewed the evidence for the idea that there may be abnormalities in
n-6 Linoleic (LA)
Gamma-linolenic
EFAs
n-3
16:2n-6
EFAs
16:3n-3
Alpha-linolenic (ALA)
delta-6-desaturase 1 16:3n-6
1 18:4rl-3
1 20:3n-6
1 20:4n-3
(GLA) Dihomogammalinolenic (DGLA)
delta-Sdesaturase Arachidonic
1 20:4n-6
1 20:5n-3
1 22:4n-6
& 22:5n-3
(AA) Adrenic
Eicosapentaenoic (EPA)
&&a-4-desaturasel 22:5n-6
22:6n-3
Docosahexaenoic
Fig. I
fatty acid and eicosanoid metabolism in various psychiatric disorders, notably schizophrenia (Feldberg, 1976; Horrobin, 1977, 1979, 1985, 1989; Horrobin et al., 1978; Van Kammen et al., 1989). It has recently been demonstrated that in the phospholipid (P), P-ethanolamine, from the frontal cortex of schizophrenics there are reduced concentrations of linoleic acid and its IS-carbon and 20carbon metabolites, balanced by a corresponding increase of the 22-carbon fatty acids (Horrobin, 1989; Horrobin et al., 1991). In plasma phospholipids, linoleic acid levels are consistently lower in schizophrenics from the same geographical area than in controls (Horrobin et al., 1989; Kaiya et al., 1991). This has been shown in Scottish. Irish, English and Japanese populations. Other fatty acid differences between normals and schizophrenics have been inconsistent. DGLA levels were significantly elevated in Japanese schizophrenics, nonsignificantly elevated in Irish schizophrenics and significantly below normal in Scottish and English schizophrenics. DGLA was found to be significantly elevated in violent criminals (Virkunnen et al., 1987). Identical twins offer unique opportunities for the study of biological markers for schizophrenia. When such twins are concordant for schizophrenia, the genome must make phenotypic schizophrenia particularly likely. When they are discordant, the effect of the genome has been over-ruled by environmental factors in one twin: whether the environment has made the normal genome schizophrenic or the schizophrenic genome normal is, of course, a matter for much debate. As part of an ongoing study of schizophrenia conducted at the National Institute of Mental Health, Washington, plasma
samples were collected from 28 pairs of identical twins. Ten of those pairs were concordant for schizophrenia, with both twins exhibiting clinical features of the disease. 18 pairs were discordant, with one exhibiting schizophrenia and the other not. The first report from this study has already appeared (Suddath et al., 1990). It demonstrated that in the pairs discordant for schizophrenia, the schizophrenic twins consistently had slightly enlarged cerebral ventricles and abnormalities in the hippocampal region. We now report on the fatty acid patterns in the plasma phospholipids and cholesterol esters in these patients.
PATIENTS
AND
METHODS
The design of the twin study, the diagnostic criteria, and information about the patients have been described in detail elsewhere (Suddath et al., 1990). In brief the diagnosis was based on revised DSM-III criteria, and in all the discordant pairs the minimum duration of discordance was 4 years. At the time of sampling, most of the schizophrenics were also being treated with neuroleptics. The characteristics of the concordant and discordant pairs are shown in Table 1. Of the ten concordant pairs, five were female as were two of the 18 discordant pairs. We thank Dr. E. Fuller Torrey of the National Institute of Mental Health, St. Elizabeth’s Hospital, Washington. DC 20032, for providing samples. Plasma samples were frozen, coded and sent by air to the Efamol Research Institute. The biochemists at the Institute were entirely unaware as to which samples came from which group. On completion of the analyses the results were sent to Dr. Bates and only then was the code broken. The samples were analysed for their fatty acid composition as described previously (Manku et al., 1983). In brief, lipids were extracted and separated into the various fractions by thin layer chromatography. The phospholipid and cholesterol ester bands were scraped off, transmethylated and the fatty acid composition of the methyl ester derivatives determined by gas chromatography. Each fatty acid in each patient group was compared statistically with the levels of that fatty acid in the normal twins. This comparison was usually
3
by Student’s t test but when fatty acid levels were below 1 mg/lOO mg and not symmetrically distributed. Mann-Whitney non-parametric tests were used.
RESULTS
The results of the phospholipid fatty acid analyses are shown in Table 2. The main features to which we wish to draw attention are the following: (1) linoleic acid levels in the discordant pairs were similar in both normals and schizophrenics but significantly below the discordant normals in the concordant schizophrenics; (2) y-linolenic and dihomo-y-linolenic acid levels were also similar in both normals and schizophrenics in the discordant pairs, but elevated in the concordant schizophrenics; (3) the most abnormal fatty acids were the 22 carbon n-6 compounds which were significantly elevated in both schizophrenic groups. However, the discordant schizophrenic twins were about half way between the concordant twins and the normals; (4) the n-3 fatty acids were normal in the discordant schizophrenics except for docosahexaenoic acid which was marginally low. Both r-linolenic and eicosapentaenoic acids were slightly elevated in the concordant schizophrenics with docosahexaenoic acid again marginally low; (5) other fatty acids were normal or near normal with a slight tendency for stearic acid to be elevated in both schizophrenic groups. The results in the cholesterol ester fraction are shown in Table 3. There were few significant abnormalities although there were some trends and differences of borderline significance (2~ < 0.1) almost entirely in the concordant schizophrenic group. In this group linoleic acid was again low while y-linolenic and dihomo-y-linolenic acids were slightly high. Oleic acid was elevated in the concordant schizophrenics.
DISCUSSION
Overall, the results support the view that there are abnormalities of fatty acid metabolism in schizo-
phrenics. This may prove to be important in view of the important roles played by those fatty acids in modulating neuronal function and in neuronal structure. However, as in all similar studies, one must insert the caveat that we cannot be certain whether some or all changes noted might be attributable either to prolonged drug treatment or to changes in lifestyle occurring because of the illness. The changes might also be attributable to some form of biochemical compensatory response to the illness. Ultimately these issues will be addressed only when the effect on the illness is investigated by manoeuvres which restore the fatty acid abnormalities to or towards normal. If such normalisation improves the clinical condition than the fatty acids are likely to be playing some causal role: if such intervention has no effect then the fatty acid changes are likely to be secondary. There were only a few differences between the discordant schizophrenics and their normal twins and these differences were largely confined to the 22-carbon fatty acids which play important roles in brain structure. The two fatty acids which were found to be elevated in the plasma phospholipids are the two whose levels were raised in frontal cortex p-ethanolamine (Horrobin, 1989; Horrobin et al., 1991). This raises the possibility of an abnormality in the conversion of precursors to the 22 carbon derivatives and in their transfer from plasma to cerebral tissue. Enzymes which might be involved are those producing elongation and desaturation of fatty acids and those incorporating fatty acids into membrane phospholipids and other complex lipids. Because they are relatively minor components of the plasma phospholipids, little attention was paid to adrenic and docosapentaenoic acids in previous studies (Horrobin et al., 1989; Kaiya et al., 1991). However, the findings in the present study are consistent with the trends in the previous investigations. Most of the fatty acid levels in the concordant schizophrenics were different from those in both the discordant schizophrenics and the normals. The same 22-carbon n-6 fatty acids were elevated as in the discordant schizophrenics, but the degree of abnormality was considerably greater. Linoleic acid was significantly reduced, consistent with all the reports to date on plasma from Japanese,
4 TABLE
1
Characteristics
of the patients
whose blood samples were investigated
in this study
No.
Age
Sex
Race
111.01 112.01 113.01 115.01 117.01 118.01
44 28 30 35 36 26
M M F F M F
Cauc. Cauc. Cauc. Cauc. Cauc. Cauc.
119.02
27
F
Cauc.
9
120.01
24
M
Cauc
2
122.01 125.02
34 28
M F
Cauc. Cauc.
20 9
301.01
31
M
African
Amer.
n/a”
301.02
31
M
African
Amer.
n/a
302.01 303.01
38 24
M M
Cauc. Cauc.
nja
303.02
24
M
Cauc.
n/a
305.01
25
M
Cauc.
da
305.02
25
M
Cauc.
n/a
308.01
41
F
Cauc.
n/a
308.02
41
F
Cauc
n/a
309.01
31
M
Cauc.
Years discordant
n/a
Medications
Mellaril 100 mg/day Prolixin D 37.5 mg/q 2 weeks Off all medication for 7 weeks Mellaril 250 mg/day Haldol 80 mg/day Lithium 1200 mg/day Tegretol 1200 mg/day Haldol 60 mg/day Tofranil 100 mg/day Navane 10 mg/day Desipramine 100 mg/day Off all medications for k 3 year Haldol 25 mg/day Haldol 50 mg. i.m. q 2 weeks Cogentin 2 mg/day Ativan PRN Prolixin (unk. dosage) Lithium (unk. dosage) Ascendin 200 mg/ IO mg Cogentin 600 mg/p.r.n. Navane IO-20 mg/day Ascendin (unk. dosage) Cogentin (unk. dosage) Prolixin 2.5 mg/day Loxitane 75 mg/day Artane 10 mg/day Lithium 1200 mg/day Azulfidine (unk. dosage) Cogentin 2 mg/day Stelazine 20 mg/day Sulfadine 4000 mgiday Tegretol 1200 mg/day Prednisone 40 mg/day Trilafon 12 mg/day Rivitral 0.5 mg/b.i.d. Trilafon 36 mg/day Kendrin 5 mg/t.i.d. Nazinan 37.5 mg/day Haldol 50 mg/i.m. q 4 weeks Haldol 80 mg/day Lithium 1200 mg/day Artane (unk. dosage) Loxitane 100 mg/day Artane (unk. dosage) Haldol 15 mg/day Benadryl (unk. dosage) Lithium 1200 mg Surmentil 200 mg/h.s. Navane 400 mg/day Mellaril 25 mg/q.i.d. p.r.n. L-tryptophan 500 mg/day Amatadine 100 mg/q.i.d.
TABLE
1 (continued)
sex
No.
YlTUS
RUCt?
Medications
discordant Cauc.
n/a
27
Cauc.
n/a
312.02
21
Cauc.
nia
313.01 313.02 314.01 314.02
26 26 29 29
Cauc. Cauc. Cauc. Cauc.
n/a n/a n/a
316.01
22
M
Cauc.
n/a
316.02
22
M
Cauc.
n/a
309.02
31
312.01
“Non-applicable TABLE
M
n/a
Stelazine 20 mg/day Lithium 900 k 122 mg on alt. days Klonopin 0.25 mg/t.i.d. Artane 5 mg/day Stelazine 20 mgiday Cogentin 2 mgiday Stelazine 20 mg/day Cogentin 2 mg!day Thorazine 400 mg/day Unknown Unknown Trilafon 16 mg/day Lithium 1500 mg/‘day Benadryl 50 mg/t.i.d. Stelazine 25 mg/day Artane 4 mg/day Mellarin 150 mg/day
since those twins were concordant
2
Concentrations of fatty acids (mg/lOO schizophrenic twins of the Jirst group. together Results are expressed as means&SD. concentrations of the fatty acids were Whitney ranking test.
mg total lipid) in the plasma phospholipid fraction of the normal monozygotic iwins. oJ the qf‘ the schizophrenic twins who were both schkophrenic and of all ~hr schtophrenics grouped
Fatty acid
Normal,
18 : 2n-6, linoleic 18 : 3n-6, y-linolenic 20 : 3n-6, dihomo-y-linolenic 20 : 4n-6, arachidonic 22 : 4n-6, adrenic 22 : 5n-6, docosapentaenoic 18 : 3n-3, cc-linolenic 20 : 5n-3, eicosapentaenoic 22 : 5n-3, docosapentaenoic 22 : 6n-3, docosahexaenoic 16 : 0, palmitic 18 : 0, stearic 18: I, oleic
30.4k4.5 0.07+0.15 2.8kO.8 12.7k2.7 0.14*0.17 0.17*0.18 0.18kO.20 0.95&0.51 1.1+1.7 4.0 + 2.6 25.6k2.2 8.1kl.3 I l.Ok2.3
Sta/istical ‘2p
