- Email: [email protected]
Increased insulin levels after OGTT load in peripheral blood and cerebrospinal fluid of patients with dementia of Alzheimer type
BIOL PSYCHIATRY 1991~O:t219-t228
1219
Increased Insulin Levels after OGTT Load in Peripheral Blood and Cerebrospinal .Fluid of Patients with Dementia of Alzheimer Type Yoshikatsu Fujisawa, Ken Sasaki, and Kazufumi Akiyama
An investigation was carried out to determine whether a disturbance of glycometabolism is associated with dementia of Alzheimer type (AD~. The first part of the s ~ was carried out on 108 AD patients and 57 normal controls (NCs). Neither the plasma level of insulin nor that of glucose differed significantly between the two groups before the oral glucose tolerance test (OGTT), wherc,,~ ~here was a significantly greater increase of the plasma insulin in the AD group than in the NC group after the OGTT. The second part of the study was carried out on 54 AD patients, 44 patients with vascular dementia (VD), and 26 NCs. Early in the morning after overnight fasting, there was no significant difference in the fasting plasma level of either glucose or insulin, or the cerebrospinal fluid (CSF) level of glucose, among the three groups, However, the CSF level of insulin was significantly higher in the AD group than in the other two groups. These results suggest that an abnormally high level of insulin, not only in the peripheral blood after OGTT load but also in the CSF after fasting, nu~ be associated with the pathology of AD.
Introduction Efforts to elucidate the etiology of dementia of Alzheimer type (AD) have expanded from neuropathological studies of the brain to include the fields of neurochemistry and, more recently, molecular biology. All these approaches, however, have focused on the brain itself, and little work has yet been done from a systemic viewpoint. Thus, apart from changes in the brain, few, if any, traits that characterize AD as a systemic disease have been highlighted. Under a hypothesis ~at a disturbance of glycometabolism might account for at least part of the pathology of AD, we undertook the present study of measuring insulin levels in peripheral blood and cerebrospinal fluid (CSF) in relatively large populations.
From Kinoko Espoir Hospital, 2908 Higashioudo, Kasaoka 714 (YF, KS) and Department of N e u r o p s y ~ , Okayama University Medical School, 2-5-1, Shikatacho, Okayama 700 (KA) Japan. Address reprint reguests to Yoshikatsu Fujisawa, M.D., Kinoko Espoir Hospital, 2908 Higashioudo, Kasaoka 714, Sapan. Received October 2, 1990; revised July 15, 1991.
© 1991 Society of Biological Psychiatry
0006-3223/9I/$03.50
1220
BIOLPSYCtH.~TaY 1991;30:1219-1228
Y. Fujisawa et al
Materials and Methods Plasma Levels of Glucose and Insulin after Glucose Subjects in study I consisted of 108 AD patients admitted to Kinoko Espoir Hospital or treated at its outpatient clinic and 57 normal elderly controls (NCs). The NC.s comprised 42 relatives of AD patients and 15 healthy persons accommodated at a home for the elderly. None of the NC persons ~ less than 25 points on MMS scores. (Anthony et al 1982; Folstein 1975). None of the controls suffered from significant medical illness. Prior to admission to the study, informed consent was obtained from the patients or their families. AD was diagnosed a~ording to DSM-HI-R criteria (American Psychiatric Association 1987) for dementia, the criteria of the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS)--Alzheimer's Disease and Relative Disorders Association (ADRDA) work group (McKharm et al 1984) for probable AD, and the magnitude of cortical atrophy in CT scan. They were also rated according to Hachinski's scoring system for cerebral ischemia (Hackinski et al 1975), which ruled out the possibility of vascular dementia (VD). None of the NCs met these criteria for the diagnosis of AD. The mean ages (plus or minus s ~ deviation) of the subjects were 72.6 ± 7.9 years for the AD group and 77.9 ± 7.3 years for the NC group. There was no significant difference in age dism'bution between the two groups. The mean ( ± SD) body weights of the subjects were 46.7 ± 7.3 kg for the AD group and 45.4 ± 9.1 kg for the NC group. Thus the groups also showed no significant difference in body weight. The AD patients and 15 NCs in the home for the elderly were maintained on 1800 k calories/day. Precise information about caloric consumption or exercise for the other controls was not obtained because they stayed at home. However, considering the similarity in body weights between the AD and NC groups, it is unlikely that the subjects in the two groups differed significantly in their caloric consumption. After supper on the day before study, all subjects fasted, and a 75-g oral glucose tolerance test (OGTI') was conducted the following morning. Blood was collected before and 30, 60, and 120 rain after the OGTI'. Plasma glucose and insulin were determined by the glucose oxidase method and radioimmunoassay, respectively. Data were analyzed by using Student's t-test for significance of differences between the AD and NC groups.
Fasting Plasma and CSF Levels of Glucose and Insulin omuy n was carried out on 34 AU patients, 44 v u patients, and 26 controls. All AD patients were admitted to Kinoko Espoir Hospital, and AD was diagnosed according to the criteria described for study I. All VD patients were admitted to Kinoko Espoir Hospital, and VD was diagnosed according to DSM-m-R criteria (American Psychiatric Association 1987) for dementia, Hachinski's scoring system for cerebral ischemia (Hachinski et al 1975), the 1986 diagnostic criteria of the Japanese Ministry of Education project team (Sekimoto 1986), and brain CT findings such as multiple small low-density areas. The controls consisted of 11 relatives of patients and 15 healthy persons accommodated at the home for the elderly. None of the NC persons had less than 25 points on MMS scores, (Anthony et al 1982; Folstein et al 1975). None of the controls suffered from physical complications or met the criteria for diagnosis of AD or VD. Prior to admission to the study, reformed consent was granted by the patients or their families. The mean (_+ SD) age of the subjects was 71.4 _+ 8.5 years for the AD group, 75.1 _+ 5.9 years for the
Insulin in Dementia of Alzeheimer Type
e[OL~¢mATR¥ 1991;30:1219-122S
1221
VD group, and 72.2 _ 8.4 years for the NC group, there being no significant difference in age distribution among the three groups. The mean (+_ SD) body weights of the subjects were 46.4 __ 6.9 kg for the AD group, 47.5 ± 8.7 kg for the VD ~ , 46.9 __ 8.4 kg for the NC group. Thus there was also no significant difference in body weight among the three groups. The AD and VD patients and 15 controls from the for the elderly were maintained on 1800 k calories/day. Although pfeci~ information about caloric consumption or exercise for the other co~U,ols was not ~ becau_~ they stayed at home, the similar body weight among the three groups suggests ~ consumption did not differ significantly. Between supper on the day before the ~ 9:00 A.M. the following morning, all subjects fasted. Then a lumbar puncture and blood collection were performed almost simultaneously. The plasma and CSF levels of glucose and insulin were determined in the same way as in the study I. Data were ~ by using one-way analysis of variance (ANOVA) for significance of difference among the three groups, and lower-order assessments were made by Student's t-test. Relafi~hips between plasma insulin and CSF insulin were examined by using S ~ ' s ran.k-order correlation coefficients.
Results Plasma Levels of Glucose and Insulin after Glucose Load There was no significant difference in the fasting plasma level of either glucose or insulin between the AD and NC groups. However, 30, 60, and 120 min after the OGTI', there was a signlicantly greater increase in the plasma level of insulin in the AD group than in the NC group (Table 1). Furthermore, the insulinogenic index, which was defined as the rate of increase in insulin at 30 min over preload value divided by the rate of increase in glucose at 30 min over preload value, was significantly higher in the AD group than in the NC group (Table 1). The plasma level of glucose, on the other hand, was significantly lower in the AD group than in the NC group only at 120 rain after the OGIT. Two patients, who had initial plasma glucose levels of 85 and 105 mg/dl, but were diagnosed as having diabetes mellitus according to the WHO criteria (1980), were included in the AD group but not in the NC group. However, these patients subsequently exhibited adequate increases in insulin levels following the OGTI" (Figure 1).
Fasting Plasma and CSF Levels of Glucose and Insulin There was no significant difference in the fasting plasma level of either glucose or insulin, or in the CSF level of glucose among the AD, VD, and NC groups (Table 2). However, the CSF level of insulin was significantly higher (p < 0.001) in the AD group (4.56 _+ 0.94 t~U/ml) than in the VD (3.50 _+ 0.29 IxU/ml) and NC groups (3.09 _+ 0.53 ttU/ml). The difference in the CSF level of insulin between the VD and NC groups was also significant (p < 0.01). The CSF-to-plasma ratio of insulin under fasting condition was 0.64 _+ 0.14 for the AD group, which was significantly (p < 0.001) higher than the values of 0.46 _+ 0.11 for the VD group and 0.43 _+ 0.10 for the NC group. There were significant positive correlations between plasma insulin and CSF insulin in the three groups (Figure 2).
1222
Y. Fujisaw~ et al
nlOL PSYCHIATRY 1991;30:1219--1228
Table 1. Plasma Levels of Glucose and Insulin after 75-g OGTr in the AD and NC Groups" Study Group
AD
N
Preload
30 min
60 rain
120 rain
0--120 min e
Glucose (mg/di) 85.2 -*" 10.3
134.5 - 29.7
124.3 - 45.1
99.0 "*" 37.7 d
101.4 "*" 83.4
0.76 -*- 0.92
108
Insulin (ttU/ml) 7.6 -4- 4.7 Glucose (mg/dl) 84.4 +_ 1 0 . 7 NC
36.5 ± 25.1~
140.4 -+ 28.5
41.3 .4- 30.6*
136.1 - 38.1
31.8 4. 26.'/*
118.1 - 30.9
86.3 -~ 56.1
141.6 ~ 77.8 0.30 _ 0.25
57
Insulin (ttU/ml) 6.4 _ 3.9 A D < NC (Glucose) AD > NC (Insulin)
Insulin°g enicc Index
18.7 -
11.6
20.2 +_ 12.9
NS
NS
NS
NS
p < 0.001
p < 0.001
19.0 -+ 11.8
39.5 "4" 29.1
p < 0.01
p < 0.01
p < 0.001
p < 0.001
p < 0.001
°Values are given as mean +- SD. bThe 0-120-min values were calculated as [(30 rain+ 60 rain + 120 rain) - l~eload value × 3l. 'The insulinogenic index was defined as the rate of increase in insulin at 30 rain over Ireload value divided by the rate of inckease in glucose at 30 min over preload value. ~Significanfly lower than NC group (p < 0.0l, Student's t-test). ~Significantly higher than NC group (/7 < 0.001, Student's t-test).
Discussion The data obtained in the present study can be summarized as follows: (1) AD patients exhibited a higher level of plasma insulin after the OGTr load; (2) the CSF level of insulin under fasting conditions was significantly higher in the AD group than in the VD and NC groups; and (3) the CSF-to-plasma ratio of insulin under fasting conditions was significantly higher in the AD group than in the NC and VD groups. Bucht et al (1983) performed a 50-g OGTT in ten AD patients and ten normal elderly persons and found that the increase in plasma insulin was higher in the former group of subjects. They also reported that, among their series of 839 patients with dementia, none of those diagnosed as having diabetes mellitus or showing a diabetic patttern following the OGTT were AD patients. Our results obtained in study I are consistent with the report of Bucht et al (1983) in that AD patients showed a greater increase in plasma insulin after the OGTI'. Among our AD patients, two were judged to show a diabetic-type blood glucose curve after the 75-g OGTr. However, even these patients showed a satisfactory elevation of insulin, indicating that they did not have true diabetes mellims, which is characterized by lowered secretion of insulin. This seems to be in agreement with the repo~ of Bucht et al (1983). Sulkava (1982) also suggested that there was a kind of dissociation between AD and diabetes mellitus. It has been demonstrated that plasma levels of insulin after fasting reflect basal secretion, and are relatively stable regardless of differences among individuals with regard to sex and daily caloric intake (Reiser et al 1985; Kaneko and Yaga 1985). Therefore, even if caloric intake in AD patients deviated from that in controls, it is conceivable that
nmL ~¥Ct~::ZA'rRY [223
Insulin in Dementia of Al~heimer Type
199It30t 12H~-|228
Insulin
Glucose
200-
n
150. .e
o°
Q•
100
100-
,' :
0-
Figure 1. Plasma glucose ~ insulin curves after 75-g OGTT in two patients diagnosed as having diabetes meHims in the AD group. Both subjects A ~ B showed a ~ c pattern of blood glucose, but insulin secretion was satisfactory.
oo.X e°°°°°°° o
xoi,,. -°x ° •
:
=-:...-~
mg/d~ 0
~/~ i
I
120
o .....
Insulin , Glucose
rain
• A Person x B Person
the higher elevation of plasma insulin in AD patients after O G ~ under fasting condition was due to AD per se. The most noteworthy finding obtained in the present study was the significant elevation in the CSF level of insulin after fasting in the AD patients as compared with the VD and NC groups. This is the first report to describe such an elevation in the CSF level of insulin in AD patients. Because the AD and VD groups did not differ significantly in caloric intake (approximately 1800 k calories/day) or body weight, it can be clearly concluded that the significantly higher level of CSF insulin is a pathophysiological feature of AD per se, and is not due to nutritional factors. There appear to be four important functions of insulin in the brain--glucose ut'dization, neuronal transmission, central nervous system (CNS) development, and feeding and body weight (Baskin et al 1987). It is still under discussion whether insulin in the brain and CSF is synthesized in CNS or transported there from the peripheral blood. Early studies showed that the concentration of insulin in the brain far exceeded that in plasma (I-Iavrankova et al 1978), and that the brain insulin content did not change if the peripheral
~224
BIOLPSYCHIATRY
Y. Fujisawa et al
1991;30:g219-122R
Table 2. Fa~tingPlasma and CSF Level of Glucose and Insulin and CSF-Plasma Ratios of Glucose and Insulin° Study Group
N
AD VD NC AD > VD AD > NC VD > NC
54 44 26
CSF glucose mg/dl 58.6 .4- 9.3 58.3 .4- 10.5 54.5 .4- 5.2 NS NS NS
Plasma Glucose mg/dl 91.0 .4- 1.7 95.1 .4- 12.6 89.5 - 7.2 NS NS NS
CSF/plasma Ratio of Glucose 0.635 .4- 0.10 0,.617 .4. 0,10 0,619 .4- 0.05 NS NS NS
CSF Insulin
Plasma Insulin
CSF~asma Ratio of Inmlin
ttU/ml 4,56 -*- 0,94 b 3,50 .4- 0.29* 3.09 .4- 0,.53 p < 0.001 p < 0.001 p < 0.01
gU/ml 7.26 .4- 1,62 8,09 .4- 3,03 7,30 .4- !.37 NS NS NS
0.64 .4- 0.14 b 0.46 ~ 0.11 0.43 __ 0.10 p < 0,001 p < 0.001 NS
"Values are given as mean ± SD. Data were assessed by one-way analysis of variance for the significance of differences among the three groups and lower-order analysis was .lone by Student's t-test. There was no significant difference in fasting plasma level of insulin among the three groups. bSignificantly higher than VD and NC groups (p < 0.001). ~Significantly higher than NC group (p < 0.01).
level of insulin was greatly altered (Havrankova et al 1981). A subsequent study using in sire hybridization with a DNA probe demonstrated that this ubiquitous presence of insulin may be due to synthesis of insulin by cells in the periventricular nucleus of the hypothalamus, and subsequent transport via the CSF to other regions of the brain (Young 1986). In contrast, Eng and Yalow (1981) reported that the concenwation of insulin in the brain was less than 20% of that in plasma, and suggested that insulin is not synthesized in the brain, but is transported from peripheral blood. In support of this view, Giddings et al (1985) suggested that immunoassayable insulin detected in extrapancreatic tissues including the brain of the adult rat is synthesized by the pancreas. It is demonstrated that brain insulin originates from peripheral blood via receptor-mediated transport through microvessels of the blood-brain barrier (Partridge et al 1985; Duffy and Partridge 1987). Van Houten et al (1983) showed that plasma insulin binds to receptors on neuronal terminals in the median eminence, where bloed-brain barrier scarcely exists, and is transported to sites deeper withi~ the hypothalamus. There are several possible reasons for elevated CSF level of insulin in the AD group. First, increased synthesis of insulin in the brain of AD patients may he the nrimarv ro~v,nn for its elevation in the CSF. Reiser et al (1985) reported that the CSF level of insulin stayed within a fixed range regardless of the plasma level of insulin, food intake, or sex. Furthermore, it was reported that short-term elevation of peripheral insulin within a physiological range of concentrations did not affect the CSF level of insulin (Ono et al 1983; Martin et al 1987). These studies indicate that insulin in the CSF arises from the CNS. Second, insulin in the bloodstream may pass into the brain. Thus, in contrast with the study of Ono et al (1983) and Martin et al (1987), it has been reported that CSF levels of insulin are elevated following intravenous injection of insulin both in the dog (Woods et al 1985) and in humans (Wallum et al 1987). Partridge et al (1985) reported that endothelial cells of the microvessels in the blood-brain barrier may bind to insulin and internalize it. The blood-brain barrier, which would otherwise regulate insulin transport to the brain, might be impaired in AD patients, as suggested by Wisniewski and Kozlowsld (1982), Elovaara et al (1985), and Scheibel and Duong (1988). Thus transport of high
2_S~
5.4
,e,
6.3
7.2
,
®
@
@
@ ®
@
@
,
,
NC G r o u p
,
9.0
insulin
8.1
.
.
6.12
6.68
~
LI..
2.68
3.04
3.40
3.9'2
4.36
" - 4.68
C
5.68 " = 5.24
9.9 (j~U/ml)
,
7s = 0.404 P < 0.05 (Spearman's correlation)
plasma
,
@
2_82 -
?_66 -I) I) @
@
@
2.98 -
@
@@ @
3.i~ --
3.30
•
@
3.9
~U/m~)
,
5.5
,
0
®
@
@
®
7.1
,
.
~
,
Group
.
10.3
insulin
8.7
.
,
D
3.96 be
4.22 -
4.48 -
4.74
m
": 5
2_40 -,
2_66
@ @
0
® O~ 0
@ ®
®
@ o, 9.2
® ® •
3.18 , , 0
" - 3.70 -
r-
:
c
11.9 (jJU/n~)
,
01's : 0.440 P < 0.01 (Spearman's correlation)
plasma
,
®
@
@
@ @
@ ®
®®
@@~) @ @®@ @
@@@@ @
$,t
@
(~/m~)
,
@
,
21.8 insulin VD Group
plasma
17.6
,
26
,
(..,AJ/n~)
7s = 0.281 P < 0.05 (Spearman's correlation) 13.4
@
@
@
Figure 2. Correlation between plasma and CSF levels of insulin after fasting. There were significant positive correlations between plasma insulin and CSF insulin in the NC, AD, and VD groups.
U. (,')
"-
r/J r-
® C = 3.48 -
3.62
3.78
3.94
4.10
(~/~)
~,~
t~ b.) ~t
.<
=-;j
-o
"m
mj
j,,~.
O
mm.
meo
~,.mo
=
1226
BIOL PSYCHIATRY 1991;30:1219-1228
Y. Fujisawa et al
concentrations of plasma insulin through the blood-brain barrier under conditions of enhanced permeability would allow insulin to pass into the CSF. Several lines of evidence indicate that insulin exerts neurotrophic effects (Aizenman and de Vellis 1987; Carofalo and Rosen 1989). A ~ n m a n and de Velfis (1987) found that insulin is important for general neuronal growth. Garofalo and Rosen (1989) recently reported that insulin plays a pivotal role in neuronal proliferation in the rat fetus. Although insulin receptors have been identified in the CNS of adult rats, especialy in the hippocampus (Unger et al 1989), it remains u n c e ~ if they n~ediate the trophic effects in the adult brain. Further studies are necessary to elucidate function of insulin receptors in the brain of the adult or aged. Recent neuropathological studies indicate that there exists sprouting of dendrites of hippocampal neurons in AD patients fl~robst et al 1983: lhara 1988). These studies raise a possibility that the pathophysiology underlying AD involves neuronal regeneration or plastic';ty as well as neuronal degeneration. Taken together, insulin may play an important role in the hippocampus, which has the highest density of insulin receptors (Unger et al 1989), and where such regenerative alterations were observed in AD. Although the functional significance of the relationship between increase in CSF level of insulin and plasma level of insulin is presently unknown, the latter fmding, as shown in study I, may indicate that AD is not only a degenerative disorder of the brain, but also is accompanied by certain ~lterations of peripheral tissues. For example, cultured skin fibroblasts from AD patients show changes in calcium homeostasis (Peterson et al 1986) and altered metabolic properties (Sims et al 1987). Thus, AD appears to be a systemic disorder in which only the most prominent pathological features are manifested in the brain. In conclusion, the present study has demonstrated a significantly greater increase of plasma insulin in AD patients as compared with normal controls after OGTI' load. It has also been shown for the first time that fasting CSF insulin levels are significantly higher in patients with AD than in those with VD or controls. Further studies will be necessary to investigate the role of insulin in the CNS of AD patients. It is hoped that our studies will stimulate further etiopathological investigations of AD as a systemic disturbance, without limitation to the brain alone. The authors w i s h to thank P r o f e s s o. r. . . S. .a. b u r o Otsuki ¢Department._ _ o f NenmngvchiatrV_r_j - - - - - J • Niravarna_Nj l lnivor~i[y
Medical School) for his helpful suggestions and critical comments.
References Aizenman Y, de Vellis J (1987): Brain neurons develop in a serum and glial free environment: Effects of transferrin, insulin, insulin-like growth factor-I and thyroid hormone on neuronal survival, growth, and differentiation. Brain Res 406:32-42. American Psychiatric Associa~on (1987): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. rev. Washington, DC: American Psychiatric Press, pp ~ - 5 5 . Anthony JC, Le Resche L, Niaz U, et al (1982): Limits of the "Mini-Mental State" as a screening test for dementia and delirium among hospital patients. Psychol Med 12:397-408. Baskin DG, Figlewicz DP, Woods SC, et al (1987): Insulin in the brain. Annu Rev Physio149:335347. Bucht G, Adolfsson R, Lithner F, et al (1983): Changes in blood glucose and insulin secretion in patients with senile dementia of Alzheimer type. Acta Med Scand 213:387-392.
Brat. P S Y a ~ A ~ V
Insulin in Dementia of Alzeheimer T)l~e
1227
1991;30:,1219-1228
Duffy KR, Pardridge WM (1987): Blood-brain barrier transcytosis of insulin in developing rabbits. Brain Res 420:32-38. Elovaara I, Icen A, Palo J, et al (1985): CSF in Alzheimer's disease studies on bloed-brain barrier function and intrathecal protein synthesis. J Neurol Sci 70:73-80. Eng J, Yalow RS (1981): Evidence against extrapancreatic insulin s y n ~ i s . Proc Natl Acad Sci USA 78:4576--4578. Folstein NF, Folstein SE, McHugh PR (1975): "Mini-mental state" A practi~ method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189-198. Garofalo RS, Rosen OM (1989): Insulin and insulinfike growth factor 1 (IGF-I) receptors during central nervous system development: Expression of two immunologically distinct IGF- I receptor subunits. Mol Cell Biol 9:2806--2817. Giddings SJ, Chirgwin J, Permutt IdA (1985): Evaluation of rat insufin messenger RNA in pancreatic and extrapancreatic tissues. Diabetologia 28:343-347. Hachinski VC, Iliff LD, Zilhka E, et al (1975): Cerebral blood flow in dementia. Arch Neurol 32:632-637. Havrankova J, Schmechel D, Roth J, et al (1978): Identification of insulin in rat brain. Proc Natl Acad Sci USA 75:5737-5741. Havrankova J, Brownstein M, Roth J (1981): Insulin and insulin receptors in rodent ~ . D/abetologia 20:268--273. lhara Y (1988): Massive somatodendritic sprouting of cortical neurons in AlTheimer's disease. Brain Res 459:138-144. Kaneko T, Yaga K (1985): Insulin, glucagon, c-peptide. Clin All-Round 3 ~ s p e c ~ issue):198201, in Japanese. Martin M, Mizrahi M, Vye P, et al (1987): The influence of acute hyperins~emia on the insulinrelated material in brain, testis, fiver, and kidney. Metabolism 36:1067-1072. McKhann G, Drachman D, Folstein M, et al (1984): Clinical diagnosis of A!zheimer's ~ : Report of the NINCDS-ADRDA Work Group under the auspices of Department of H ~ t h and Human Sciences Task Force on A!Theimer's disease. Ne~rology 34:939-944. Ono T, Steffeus AB, Sasaki K (1983): Influence of peripheral and inwacerebroventricular glucose and insulin infusions on perpheral and cerebrospinal fluid glucose and i n s ~ levels. Physiol Behav 30:301-306. Pardfidge WM, Eisenberg J, Yang J (1985): Human blood-brain barrier insufin receptor. J Neurochem 44:1771-1778. Peterson C, Ratan RR, Sh]anski ML, et -,11(1986): Cytosofic free c~cium and ceil spreading decrease ,', 6h,~hl2~tc from aged and ~J~'h'~imer donor. Proc ~,,,i ac~4 Sci HCA ~a.'moo ~ Pmbst A, Basler V, Bron B, et ed (1983): Neuritic plaques in senile dementia of AIzheimer type. A Golgi analysis in the hippocampal region. Brain Res 268:249-254. Reiser M, Lenz E, Bernstein H-G, et al (1985): Insulin-like immunoreactivity in human cerebrospinal-fluid is independent of insulin blood levels. Human Neurobiol 4:53-55. Scheibel AB, Duong T (1988): On the possibte relationship of cortical microvascular pathology to blood brain barrier changes in Alzheimer's disease. Neurobiol Aging 9:41-42. Sekimoto H (1986): Diagnostic criteria for cerebrovascular disease. Jpn J CHn Med 44(special issue: Diagnostic criteria and clinical aspects of cerebrovascular disorders): 11-22, in Japanese. Sims N, Finegan JM, Blass JP (1987): Altered metabolic properties of cultured skin fibroblasts in Alzheimer's disease. Ann Neurol 21:451-457. Sulkava R (1982): Alzheimer's disease and senile dementia of A|zheimer type. Acta Neurol Scand 65:636--650. Unger J, McNeill TH, Moxley RT 11I, et al (1989): Distribution of insulin receptor-like immunoreactivity in the rat forebrain. Neuroscience 31:143--157. Van Houten M, Nance DM, Gauthier S, et al (1983): Origin of insulin-receptive nerve terminals in rat median eminence. Endocrinology 113:1393-1399. 4tatw,J i t w e ,
la~m.Jri.tJ
L
,Ig.~lHt~
~ICA
atvf.~t~l~
~a~
I.Jt.~tTg
~J,..J.$.~c~v.Jr"--lJW&
•
1228
BIOLPSYCHIATRY 1991;30:1219--1228
Y. Fujisawa et al
W ~ u m BJ, Taborsky GJ Jr, Porte D Jr, et al (1987): ~ p i n a l fluid insulin levels increase during intravenous insulin infusion in man. J Clin Endocr/no/Metab 64:190-194. WHO Expert Committee on Diabetes Mellitus (1980): Seco+~ Report, Technical Report Series 646 WHO, Geneva. Wisnieski HM, Kozlowski PB (1982): Evidence of blood-brain barrier changes in senite dementia of the Alzheimer type. Ann NY Acad Sci 396:119-129. Woods SC, Porte D Jr, Bobbioni E, et al (1985): Insulin: Its relationship to the central nervous system and to the control of food intake and body weight. Am J Clin Ntar 42:1063--1071. Young WS (1986): Periventricular hypothalamic cells in the rat brain contain insulin mRNA. Neuropeptides 8:93-97.
1219
Increased Insulin Levels after OGTT Load in Peripheral Blood and Cerebrospinal .Fluid of Patients with Dementia of Alzheimer Type Yoshikatsu Fujisawa, Ken Sasaki, and Kazufumi Akiyama
An investigation was carried out to determine whether a disturbance of glycometabolism is associated with dementia of Alzheimer type (AD~. The first part of the s ~ was carried out on 108 AD patients and 57 normal controls (NCs). Neither the plasma level of insulin nor that of glucose differed significantly between the two groups before the oral glucose tolerance test (OGTT), wherc,,~ ~here was a significantly greater increase of the plasma insulin in the AD group than in the NC group after the OGTT. The second part of the study was carried out on 54 AD patients, 44 patients with vascular dementia (VD), and 26 NCs. Early in the morning after overnight fasting, there was no significant difference in the fasting plasma level of either glucose or insulin, or the cerebrospinal fluid (CSF) level of glucose, among the three groups, However, the CSF level of insulin was significantly higher in the AD group than in the other two groups. These results suggest that an abnormally high level of insulin, not only in the peripheral blood after OGTT load but also in the CSF after fasting, nu~ be associated with the pathology of AD.
Introduction Efforts to elucidate the etiology of dementia of Alzheimer type (AD) have expanded from neuropathological studies of the brain to include the fields of neurochemistry and, more recently, molecular biology. All these approaches, however, have focused on the brain itself, and little work has yet been done from a systemic viewpoint. Thus, apart from changes in the brain, few, if any, traits that characterize AD as a systemic disease have been highlighted. Under a hypothesis ~at a disturbance of glycometabolism might account for at least part of the pathology of AD, we undertook the present study of measuring insulin levels in peripheral blood and cerebrospinal fluid (CSF) in relatively large populations.
From Kinoko Espoir Hospital, 2908 Higashioudo, Kasaoka 714 (YF, KS) and Department of N e u r o p s y ~ , Okayama University Medical School, 2-5-1, Shikatacho, Okayama 700 (KA) Japan. Address reprint reguests to Yoshikatsu Fujisawa, M.D., Kinoko Espoir Hospital, 2908 Higashioudo, Kasaoka 714, Sapan. Received October 2, 1990; revised July 15, 1991.
© 1991 Society of Biological Psychiatry
0006-3223/9I/$03.50
1220
BIOLPSYCtH.~TaY 1991;30:1219-1228
Y. Fujisawa et al
Materials and Methods Plasma Levels of Glucose and Insulin after Glucose Subjects in study I consisted of 108 AD patients admitted to Kinoko Espoir Hospital or treated at its outpatient clinic and 57 normal elderly controls (NCs). The NC.s comprised 42 relatives of AD patients and 15 healthy persons accommodated at a home for the elderly. None of the NC persons ~ less than 25 points on MMS scores. (Anthony et al 1982; Folstein 1975). None of the controls suffered from significant medical illness. Prior to admission to the study, informed consent was obtained from the patients or their families. AD was diagnosed a~ording to DSM-HI-R criteria (American Psychiatric Association 1987) for dementia, the criteria of the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS)--Alzheimer's Disease and Relative Disorders Association (ADRDA) work group (McKharm et al 1984) for probable AD, and the magnitude of cortical atrophy in CT scan. They were also rated according to Hachinski's scoring system for cerebral ischemia (Hackinski et al 1975), which ruled out the possibility of vascular dementia (VD). None of the NCs met these criteria for the diagnosis of AD. The mean ages (plus or minus s ~ deviation) of the subjects were 72.6 ± 7.9 years for the AD group and 77.9 ± 7.3 years for the NC group. There was no significant difference in age dism'bution between the two groups. The mean ( ± SD) body weights of the subjects were 46.7 ± 7.3 kg for the AD group and 45.4 ± 9.1 kg for the NC group. Thus the groups also showed no significant difference in body weight. The AD patients and 15 NCs in the home for the elderly were maintained on 1800 k calories/day. Precise information about caloric consumption or exercise for the other controls was not obtained because they stayed at home. However, considering the similarity in body weights between the AD and NC groups, it is unlikely that the subjects in the two groups differed significantly in their caloric consumption. After supper on the day before study, all subjects fasted, and a 75-g oral glucose tolerance test (OGTI') was conducted the following morning. Blood was collected before and 30, 60, and 120 rain after the OGTI'. Plasma glucose and insulin were determined by the glucose oxidase method and radioimmunoassay, respectively. Data were analyzed by using Student's t-test for significance of differences between the AD and NC groups.
Fasting Plasma and CSF Levels of Glucose and Insulin omuy n was carried out on 34 AU patients, 44 v u patients, and 26 controls. All AD patients were admitted to Kinoko Espoir Hospital, and AD was diagnosed according to the criteria described for study I. All VD patients were admitted to Kinoko Espoir Hospital, and VD was diagnosed according to DSM-m-R criteria (American Psychiatric Association 1987) for dementia, Hachinski's scoring system for cerebral ischemia (Hachinski et al 1975), the 1986 diagnostic criteria of the Japanese Ministry of Education project team (Sekimoto 1986), and brain CT findings such as multiple small low-density areas. The controls consisted of 11 relatives of patients and 15 healthy persons accommodated at the home for the elderly. None of the NC persons had less than 25 points on MMS scores, (Anthony et al 1982; Folstein et al 1975). None of the controls suffered from physical complications or met the criteria for diagnosis of AD or VD. Prior to admission to the study, reformed consent was granted by the patients or their families. The mean (_+ SD) age of the subjects was 71.4 _+ 8.5 years for the AD group, 75.1 _+ 5.9 years for the
Insulin in Dementia of Alzeheimer Type
e[OL~¢mATR¥ 1991;30:1219-122S
1221
VD group, and 72.2 _ 8.4 years for the NC group, there being no significant difference in age distribution among the three groups. The mean (+_ SD) body weights of the subjects were 46.4 __ 6.9 kg for the AD group, 47.5 ± 8.7 kg for the VD ~ , 46.9 __ 8.4 kg for the NC group. Thus there was also no significant difference in body weight among the three groups. The AD and VD patients and 15 controls from the for the elderly were maintained on 1800 k calories/day. Although pfeci~ information about caloric consumption or exercise for the other co~U,ols was not ~ becau_~ they stayed at home, the similar body weight among the three groups suggests ~ consumption did not differ significantly. Between supper on the day before the ~ 9:00 A.M. the following morning, all subjects fasted. Then a lumbar puncture and blood collection were performed almost simultaneously. The plasma and CSF levels of glucose and insulin were determined in the same way as in the study I. Data were ~ by using one-way analysis of variance (ANOVA) for significance of difference among the three groups, and lower-order assessments were made by Student's t-test. Relafi~hips between plasma insulin and CSF insulin were examined by using S ~ ' s ran.k-order correlation coefficients.
Results Plasma Levels of Glucose and Insulin after Glucose Load There was no significant difference in the fasting plasma level of either glucose or insulin between the AD and NC groups. However, 30, 60, and 120 min after the OGTI', there was a signlicantly greater increase in the plasma level of insulin in the AD group than in the NC group (Table 1). Furthermore, the insulinogenic index, which was defined as the rate of increase in insulin at 30 min over preload value divided by the rate of increase in glucose at 30 min over preload value, was significantly higher in the AD group than in the NC group (Table 1). The plasma level of glucose, on the other hand, was significantly lower in the AD group than in the NC group only at 120 rain after the OGIT. Two patients, who had initial plasma glucose levels of 85 and 105 mg/dl, but were diagnosed as having diabetes mellitus according to the WHO criteria (1980), were included in the AD group but not in the NC group. However, these patients subsequently exhibited adequate increases in insulin levels following the OGTI" (Figure 1).
Fasting Plasma and CSF Levels of Glucose and Insulin There was no significant difference in the fasting plasma level of either glucose or insulin, or in the CSF level of glucose among the AD, VD, and NC groups (Table 2). However, the CSF level of insulin was significantly higher (p < 0.001) in the AD group (4.56 _+ 0.94 t~U/ml) than in the VD (3.50 _+ 0.29 IxU/ml) and NC groups (3.09 _+ 0.53 ttU/ml). The difference in the CSF level of insulin between the VD and NC groups was also significant (p < 0.01). The CSF-to-plasma ratio of insulin under fasting condition was 0.64 _+ 0.14 for the AD group, which was significantly (p < 0.001) higher than the values of 0.46 _+ 0.11 for the VD group and 0.43 _+ 0.10 for the NC group. There were significant positive correlations between plasma insulin and CSF insulin in the three groups (Figure 2).
1222
Y. Fujisaw~ et al
nlOL PSYCHIATRY 1991;30:1219--1228
Table 1. Plasma Levels of Glucose and Insulin after 75-g OGTr in the AD and NC Groups" Study Group
AD
N
Preload
30 min
60 rain
120 rain
0--120 min e
Glucose (mg/di) 85.2 -*" 10.3
134.5 - 29.7
124.3 - 45.1
99.0 "*" 37.7 d
101.4 "*" 83.4
0.76 -*- 0.92
108
Insulin (ttU/ml) 7.6 -4- 4.7 Glucose (mg/dl) 84.4 +_ 1 0 . 7 NC
36.5 ± 25.1~
140.4 -+ 28.5
41.3 .4- 30.6*
136.1 - 38.1
31.8 4. 26.'/*
118.1 - 30.9
86.3 -~ 56.1
141.6 ~ 77.8 0.30 _ 0.25
57
Insulin (ttU/ml) 6.4 _ 3.9 A D < NC (Glucose) AD > NC (Insulin)
Insulin°g enicc Index
18.7 -
11.6
20.2 +_ 12.9
NS
NS
NS
NS
p < 0.001
p < 0.001
19.0 -+ 11.8
39.5 "4" 29.1
p < 0.01
p < 0.01
p < 0.001
p < 0.001
p < 0.001
°Values are given as mean +- SD. bThe 0-120-min values were calculated as [(30 rain+ 60 rain + 120 rain) - l~eload value × 3l. 'The insulinogenic index was defined as the rate of increase in insulin at 30 rain over Ireload value divided by the rate of inckease in glucose at 30 min over preload value. ~Significanfly lower than NC group (p < 0.0l, Student's t-test). ~Significantly higher than NC group (/7 < 0.001, Student's t-test).
Discussion The data obtained in the present study can be summarized as follows: (1) AD patients exhibited a higher level of plasma insulin after the OGTr load; (2) the CSF level of insulin under fasting conditions was significantly higher in the AD group than in the VD and NC groups; and (3) the CSF-to-plasma ratio of insulin under fasting conditions was significantly higher in the AD group than in the NC and VD groups. Bucht et al (1983) performed a 50-g OGTT in ten AD patients and ten normal elderly persons and found that the increase in plasma insulin was higher in the former group of subjects. They also reported that, among their series of 839 patients with dementia, none of those diagnosed as having diabetes mellitus or showing a diabetic patttern following the OGTT were AD patients. Our results obtained in study I are consistent with the report of Bucht et al (1983) in that AD patients showed a greater increase in plasma insulin after the OGTI'. Among our AD patients, two were judged to show a diabetic-type blood glucose curve after the 75-g OGTr. However, even these patients showed a satisfactory elevation of insulin, indicating that they did not have true diabetes mellims, which is characterized by lowered secretion of insulin. This seems to be in agreement with the repo~ of Bucht et al (1983). Sulkava (1982) also suggested that there was a kind of dissociation between AD and diabetes mellitus. It has been demonstrated that plasma levels of insulin after fasting reflect basal secretion, and are relatively stable regardless of differences among individuals with regard to sex and daily caloric intake (Reiser et al 1985; Kaneko and Yaga 1985). Therefore, even if caloric intake in AD patients deviated from that in controls, it is conceivable that
nmL ~¥Ct~::ZA'rRY [223
Insulin in Dementia of Al~heimer Type
199It30t 12H~-|228
Insulin
Glucose
200-
n
150. .e
o°
Q•
100
100-
,' :
0-
Figure 1. Plasma glucose ~ insulin curves after 75-g OGTT in two patients diagnosed as having diabetes meHims in the AD group. Both subjects A ~ B showed a ~ c pattern of blood glucose, but insulin secretion was satisfactory.
oo.X e°°°°°°° o
xoi,,. -°x ° •
:
=-:...-~
mg/d~ 0
~/~ i
I
120
o .....
Insulin , Glucose
rain
• A Person x B Person
the higher elevation of plasma insulin in AD patients after O G ~ under fasting condition was due to AD per se. The most noteworthy finding obtained in the present study was the significant elevation in the CSF level of insulin after fasting in the AD patients as compared with the VD and NC groups. This is the first report to describe such an elevation in the CSF level of insulin in AD patients. Because the AD and VD groups did not differ significantly in caloric intake (approximately 1800 k calories/day) or body weight, it can be clearly concluded that the significantly higher level of CSF insulin is a pathophysiological feature of AD per se, and is not due to nutritional factors. There appear to be four important functions of insulin in the brain--glucose ut'dization, neuronal transmission, central nervous system (CNS) development, and feeding and body weight (Baskin et al 1987). It is still under discussion whether insulin in the brain and CSF is synthesized in CNS or transported there from the peripheral blood. Early studies showed that the concentration of insulin in the brain far exceeded that in plasma (I-Iavrankova et al 1978), and that the brain insulin content did not change if the peripheral
~224
BIOLPSYCHIATRY
Y. Fujisawa et al
1991;30:g219-122R
Table 2. Fa~tingPlasma and CSF Level of Glucose and Insulin and CSF-Plasma Ratios of Glucose and Insulin° Study Group
N
AD VD NC AD > VD AD > NC VD > NC
54 44 26
CSF glucose mg/dl 58.6 .4- 9.3 58.3 .4- 10.5 54.5 .4- 5.2 NS NS NS
Plasma Glucose mg/dl 91.0 .4- 1.7 95.1 .4- 12.6 89.5 - 7.2 NS NS NS
CSF/plasma Ratio of Glucose 0.635 .4- 0.10 0,.617 .4. 0,10 0,619 .4- 0.05 NS NS NS
CSF Insulin
Plasma Insulin
CSF~asma Ratio of Inmlin
ttU/ml 4,56 -*- 0,94 b 3,50 .4- 0.29* 3.09 .4- 0,.53 p < 0.001 p < 0.001 p < 0.01
gU/ml 7.26 .4- 1,62 8,09 .4- 3,03 7,30 .4- !.37 NS NS NS
0.64 .4- 0.14 b 0.46 ~ 0.11 0.43 __ 0.10 p < 0,001 p < 0.001 NS
"Values are given as mean ± SD. Data were assessed by one-way analysis of variance for the significance of differences among the three groups and lower-order analysis was .lone by Student's t-test. There was no significant difference in fasting plasma level of insulin among the three groups. bSignificantly higher than VD and NC groups (p < 0.001). ~Significantly higher than NC group (p < 0.01).
level of insulin was greatly altered (Havrankova et al 1981). A subsequent study using in sire hybridization with a DNA probe demonstrated that this ubiquitous presence of insulin may be due to synthesis of insulin by cells in the periventricular nucleus of the hypothalamus, and subsequent transport via the CSF to other regions of the brain (Young 1986). In contrast, Eng and Yalow (1981) reported that the concenwation of insulin in the brain was less than 20% of that in plasma, and suggested that insulin is not synthesized in the brain, but is transported from peripheral blood. In support of this view, Giddings et al (1985) suggested that immunoassayable insulin detected in extrapancreatic tissues including the brain of the adult rat is synthesized by the pancreas. It is demonstrated that brain insulin originates from peripheral blood via receptor-mediated transport through microvessels of the blood-brain barrier (Partridge et al 1985; Duffy and Partridge 1987). Van Houten et al (1983) showed that plasma insulin binds to receptors on neuronal terminals in the median eminence, where bloed-brain barrier scarcely exists, and is transported to sites deeper withi~ the hypothalamus. There are several possible reasons for elevated CSF level of insulin in the AD group. First, increased synthesis of insulin in the brain of AD patients may he the nrimarv ro~v,nn for its elevation in the CSF. Reiser et al (1985) reported that the CSF level of insulin stayed within a fixed range regardless of the plasma level of insulin, food intake, or sex. Furthermore, it was reported that short-term elevation of peripheral insulin within a physiological range of concentrations did not affect the CSF level of insulin (Ono et al 1983; Martin et al 1987). These studies indicate that insulin in the CSF arises from the CNS. Second, insulin in the bloodstream may pass into the brain. Thus, in contrast with the study of Ono et al (1983) and Martin et al (1987), it has been reported that CSF levels of insulin are elevated following intravenous injection of insulin both in the dog (Woods et al 1985) and in humans (Wallum et al 1987). Partridge et al (1985) reported that endothelial cells of the microvessels in the blood-brain barrier may bind to insulin and internalize it. The blood-brain barrier, which would otherwise regulate insulin transport to the brain, might be impaired in AD patients, as suggested by Wisniewski and Kozlowsld (1982), Elovaara et al (1985), and Scheibel and Duong (1988). Thus transport of high
2_S~
5.4
,e,
6.3
7.2
,
®
@
@
@ ®
@
@
,
,
NC G r o u p
,
9.0
insulin
8.1
.
.
6.12
6.68
~
LI..
2.68
3.04
3.40
3.9'2
4.36
" - 4.68
C
5.68 " = 5.24
9.9 (j~U/ml)
,
7s = 0.404 P < 0.05 (Spearman's correlation)
plasma
,
@
2_82 -
?_66 -I) I) @
@
@
2.98 -
@
@@ @
3.i~ --
3.30
•
@
3.9
~U/m~)
,
5.5
,
0
®
@
@
®
7.1
,
.
~
,
Group
.
10.3
insulin
8.7
.
,
D
3.96 be
4.22 -
4.48 -
4.74
m
": 5
2_40 -,
2_66
@ @
0
® O~ 0
@ ®
®
@ o, 9.2
® ® •
3.18 , , 0
" - 3.70 -
r-
:
c
11.9 (jJU/n~)
,
01's : 0.440 P < 0.01 (Spearman's correlation)
plasma
,
®
@
@
@ @
@ ®
®®
@@~) @ @®@ @
@@@@ @
$,t
@
(~/m~)
,
@
,
21.8 insulin VD Group
plasma
17.6
,
26
,
(..,AJ/n~)
7s = 0.281 P < 0.05 (Spearman's correlation) 13.4
@
@
@
Figure 2. Correlation between plasma and CSF levels of insulin after fasting. There were significant positive correlations between plasma insulin and CSF insulin in the NC, AD, and VD groups.
U. (,')
"-
r/J r-
® C = 3.48 -
3.62
3.78
3.94
4.10
(~/~)
~,~
t~ b.) ~t
.<
=-;j
-o
"m
mj
j,,~.
O
mm.
meo
~,.mo
=
1226
BIOL PSYCHIATRY 1991;30:1219-1228
Y. Fujisawa et al
concentrations of plasma insulin through the blood-brain barrier under conditions of enhanced permeability would allow insulin to pass into the CSF. Several lines of evidence indicate that insulin exerts neurotrophic effects (Aizenman and de Vellis 1987; Carofalo and Rosen 1989). A ~ n m a n and de Velfis (1987) found that insulin is important for general neuronal growth. Garofalo and Rosen (1989) recently reported that insulin plays a pivotal role in neuronal proliferation in the rat fetus. Although insulin receptors have been identified in the CNS of adult rats, especialy in the hippocampus (Unger et al 1989), it remains u n c e ~ if they n~ediate the trophic effects in the adult brain. Further studies are necessary to elucidate function of insulin receptors in the brain of the adult or aged. Recent neuropathological studies indicate that there exists sprouting of dendrites of hippocampal neurons in AD patients fl~robst et al 1983: lhara 1988). These studies raise a possibility that the pathophysiology underlying AD involves neuronal regeneration or plastic';ty as well as neuronal degeneration. Taken together, insulin may play an important role in the hippocampus, which has the highest density of insulin receptors (Unger et al 1989), and where such regenerative alterations were observed in AD. Although the functional significance of the relationship between increase in CSF level of insulin and plasma level of insulin is presently unknown, the latter fmding, as shown in study I, may indicate that AD is not only a degenerative disorder of the brain, but also is accompanied by certain ~lterations of peripheral tissues. For example, cultured skin fibroblasts from AD patients show changes in calcium homeostasis (Peterson et al 1986) and altered metabolic properties (Sims et al 1987). Thus, AD appears to be a systemic disorder in which only the most prominent pathological features are manifested in the brain. In conclusion, the present study has demonstrated a significantly greater increase of plasma insulin in AD patients as compared with normal controls after OGTI' load. It has also been shown for the first time that fasting CSF insulin levels are significantly higher in patients with AD than in those with VD or controls. Further studies will be necessary to investigate the role of insulin in the CNS of AD patients. It is hoped that our studies will stimulate further etiopathological investigations of AD as a systemic disturbance, without limitation to the brain alone. The authors w i s h to thank P r o f e s s o. r. . . S. .a. b u r o Otsuki ¢Department._ _ o f NenmngvchiatrV_r_j - - - - - J • Niravarna_Nj l lnivor~i[y
Medical School) for his helpful suggestions and critical comments.
References Aizenman Y, de Vellis J (1987): Brain neurons develop in a serum and glial free environment: Effects of transferrin, insulin, insulin-like growth factor-I and thyroid hormone on neuronal survival, growth, and differentiation. Brain Res 406:32-42. American Psychiatric Associa~on (1987): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. rev. Washington, DC: American Psychiatric Press, pp ~ - 5 5 . Anthony JC, Le Resche L, Niaz U, et al (1982): Limits of the "Mini-Mental State" as a screening test for dementia and delirium among hospital patients. Psychol Med 12:397-408. Baskin DG, Figlewicz DP, Woods SC, et al (1987): Insulin in the brain. Annu Rev Physio149:335347. Bucht G, Adolfsson R, Lithner F, et al (1983): Changes in blood glucose and insulin secretion in patients with senile dementia of Alzheimer type. Acta Med Scand 213:387-392.
Brat. P S Y a ~ A ~ V
Insulin in Dementia of Alzeheimer T)l~e
1227
1991;30:,1219-1228
Duffy KR, Pardridge WM (1987): Blood-brain barrier transcytosis of insulin in developing rabbits. Brain Res 420:32-38. Elovaara I, Icen A, Palo J, et al (1985): CSF in Alzheimer's disease studies on bloed-brain barrier function and intrathecal protein synthesis. J Neurol Sci 70:73-80. Eng J, Yalow RS (1981): Evidence against extrapancreatic insulin s y n ~ i s . Proc Natl Acad Sci USA 78:4576--4578. Folstein NF, Folstein SE, McHugh PR (1975): "Mini-mental state" A practi~ method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189-198. Garofalo RS, Rosen OM (1989): Insulin and insulinfike growth factor 1 (IGF-I) receptors during central nervous system development: Expression of two immunologically distinct IGF- I receptor subunits. Mol Cell Biol 9:2806--2817. Giddings SJ, Chirgwin J, Permutt IdA (1985): Evaluation of rat insufin messenger RNA in pancreatic and extrapancreatic tissues. Diabetologia 28:343-347. Hachinski VC, Iliff LD, Zilhka E, et al (1975): Cerebral blood flow in dementia. Arch Neurol 32:632-637. Havrankova J, Schmechel D, Roth J, et al (1978): Identification of insulin in rat brain. Proc Natl Acad Sci USA 75:5737-5741. Havrankova J, Brownstein M, Roth J (1981): Insulin and insulin receptors in rodent ~ . D/abetologia 20:268--273. lhara Y (1988): Massive somatodendritic sprouting of cortical neurons in AlTheimer's disease. Brain Res 459:138-144. Kaneko T, Yaga K (1985): Insulin, glucagon, c-peptide. Clin All-Round 3 ~ s p e c ~ issue):198201, in Japanese. Martin M, Mizrahi M, Vye P, et al (1987): The influence of acute hyperins~emia on the insulinrelated material in brain, testis, fiver, and kidney. Metabolism 36:1067-1072. McKhann G, Drachman D, Folstein M, et al (1984): Clinical diagnosis of A!zheimer's ~ : Report of the NINCDS-ADRDA Work Group under the auspices of Department of H ~ t h and Human Sciences Task Force on A!Theimer's disease. Ne~rology 34:939-944. Ono T, Steffeus AB, Sasaki K (1983): Influence of peripheral and inwacerebroventricular glucose and insulin infusions on perpheral and cerebrospinal fluid glucose and i n s ~ levels. Physiol Behav 30:301-306. Pardfidge WM, Eisenberg J, Yang J (1985): Human blood-brain barrier insufin receptor. J Neurochem 44:1771-1778. Peterson C, Ratan RR, Sh]anski ML, et -,11(1986): Cytosofic free c~cium and ceil spreading decrease ,', 6h,~hl2~tc from aged and ~J~'h'~imer donor. Proc ~,,,i ac~4 Sci HCA ~a.'moo ~ Pmbst A, Basler V, Bron B, et ed (1983): Neuritic plaques in senile dementia of AIzheimer type. A Golgi analysis in the hippocampal region. Brain Res 268:249-254. Reiser M, Lenz E, Bernstein H-G, et al (1985): Insulin-like immunoreactivity in human cerebrospinal-fluid is independent of insulin blood levels. Human Neurobiol 4:53-55. Scheibel AB, Duong T (1988): On the possibte relationship of cortical microvascular pathology to blood brain barrier changes in Alzheimer's disease. Neurobiol Aging 9:41-42. Sekimoto H (1986): Diagnostic criteria for cerebrovascular disease. Jpn J CHn Med 44(special issue: Diagnostic criteria and clinical aspects of cerebrovascular disorders): 11-22, in Japanese. Sims N, Finegan JM, Blass JP (1987): Altered metabolic properties of cultured skin fibroblasts in Alzheimer's disease. Ann Neurol 21:451-457. Sulkava R (1982): Alzheimer's disease and senile dementia of A|zheimer type. Acta Neurol Scand 65:636--650. Unger J, McNeill TH, Moxley RT 11I, et al (1989): Distribution of insulin receptor-like immunoreactivity in the rat forebrain. Neuroscience 31:143--157. Van Houten M, Nance DM, Gauthier S, et al (1983): Origin of insulin-receptive nerve terminals in rat median eminence. Endocrinology 113:1393-1399. 4tatw,J i t w e ,
la~m.Jri.tJ
L
,Ig.~lHt~
~ICA
atvf.~t~l~
~a~
I.Jt.~tTg
~J,..J.$.~c~v.Jr"--lJW&
•
1228
BIOLPSYCHIATRY 1991;30:1219--1228
Y. Fujisawa et al
W ~ u m BJ, Taborsky GJ Jr, Porte D Jr, et al (1987): ~ p i n a l fluid insulin levels increase during intravenous insulin infusion in man. J Clin Endocr/no/Metab 64:190-194. WHO Expert Committee on Diabetes Mellitus (1980): Seco+~ Report, Technical Report Series 646 WHO, Geneva. Wisnieski HM, Kozlowski PB (1982): Evidence of blood-brain barrier changes in senite dementia of the Alzheimer type. Ann NY Acad Sci 396:119-129. Woods SC, Porte D Jr, Bobbioni E, et al (1985): Insulin: Its relationship to the central nervous system and to the control of food intake and body weight. Am J Clin Ntar 42:1063--1071. Young WS (1986): Periventricular hypothalamic cells in the rat brain contain insulin mRNA. Neuropeptides 8:93-97.