Reduced Gastrointestinal Absorption of Calcium in Dementia I. NICOL FERRIER, ALAN LEAKE, GEOFFREY A. TAYLOR, IAN G. McKEITH, ANDREW F. FAIRBAIRN, CHRISTOPHER J. ROBINSON, ROGER M. FRANCIS, JAMES A. EDWARDSON
Summary Several reports have suggested that the neurodegenerative change in Alzheimer-type dementia (ATD) may be related to alterations in calcium homoeostasis. The absorption of radiocalcium (4SCa) in 26 ATD subjects and 11 patients with multi-infarct dementia (MID) was compared to 24 normal age- and sexmatched controls. The absorption of radiocalcium was significantly lower in both ATD and M I D when compared to controls. The reduced 4SCa absorption in ATD occurred in the presence of normal plasma concentrations of PTH and vitamin D metabolites and the serum concentrations of calcium and aluminium were in the normal range. The data suggest that the reduced uptake of radioactive calcium observed in A T D is a non-specific derangement.
Introduction Alterations of cellular calcium homoeostasis may be involved in the chain of events that leads to neurodegenerative changes in Alzheimer's disease and recent evidence suggests that such alterations may be expressed in extraneural tissue. The evidence for this statement is diverse and complex and a summary is given below: 1. Impairment of neuronal calcium homoeostasis or of calcium-dependent processes has been implicated in degenerative disorders of the mammalian nervous system such as those associated with ageing or environmental factors [1-3] including toxin-induced nerve-cell death [4]. An elevation of cytoplasmic free calcium in neurones seems to play an important part in brain damage following ischaemic or hypoglycaemic episodes [5, 6]. In neurones at rest the intracellular concentration of free-calcium ions is over four orders of magnitude lower than
extracellular free calcium. This remarkable transmembrane gradient seems essential for maintaining cell integrity because either a decrease in extracellular calcium or an increase in intracellular calcium can lead to neuronal degeneration [7, 8]. Since the surface membrane of neurones is permeable to calcium ions, the low intracellular calcium is achieved by an efficient buffering system for free-calcium ions following their entry into the cytoplasm [9]. The elements of the intracellular calciumbuffering system include the mitochondria, the endoplasmic reticulum, calcium-binding proteins in the cytosol and ATP-linked phosphorylating processes. There is evidence that at least some of the elements of the calcium-buffering systems are affected in Alzheimer's disease. The endoplasmic reticulum associated enzyme aneutral-glucosidase is markedly reduced in Alzheimer-type dementia (ATD) in temporal cortex and hippocampus [10]. Neurones containing parvalbumin (a cytosolic calcium-bindAge and Ageing 1990,19:368-375
REDUCED ABSORPTION OF CALCIUM IN DEMENTIA
ing protein) are markedly reduced in number in ATD cortex [11]. Pyruvate dehydrogenase enzyme complex (PHDC) activity is reduced in ATD and this highly regulated enzyme appears to have an important role in neuronal calcium homoeostasis [12-14]. It is not clear at present if these changes in calcium buffering mechanisms in ATD are primary or secondary events. Aberrant phosphorylation of microtubuleassociated protein is likely to be an important mechanism in the pathogenesis of neurofibrillary tangles [15] and may be related to impaired intracellular calcium homoeostasis with consequent activation of calcium-dependent protein kinases [7]. 2. There is evidence for abnormal intracerebral accumulation of calcium in some dementing disorders. Garruto et al. [1] demonstrated increased calcium deposition in tangle-bearing neurones in the parkinsonism-dementia complex of Guam. Accumulation of aluminium within these neurones also occurs in this disorder and it has been suggested that these changes are due to secondary hyperparathyroidism resulting from an environment rich in aluminium but low in calcium and magnesium. There is evidence for increased calcium in cortex in pathologically confirmed A T D [16]. Mann [17] has shown that the prevalence and severity of calcification in the basal ganglia is greater (for age) in Down's syndrome and in patients with Alzheimer's disease under 75 years of age. However, calcification of the basal ganglia is frequently encountered in patients with a deficiency of parathormone and in such patients epilepsy, parkinsonism and dementia are often present [18]. 3. There is some evidence that alterations in cellular calcium homoeostasis may be expressed in extraneural tissue. Reduced calcium uptake has been demonstrated by cultured skin fibroblasts [19] and by mitogen-stimulated lymphocytes [20] taken from patients with Alzheimertype disease. It is noteworthy that Down's syndrome patients, most of whom develop Alzheimer-type neuropathological changes over the age of 35 [21] have reduced plasma calcium levels [22, 23] and reduced calcium uptake into platelets [24]. Senile plaques and neurofibrillary tangles also occur in normal elderly individuals. Cal-
369
cium absorption from the bowel declines with advancing age [25] and this may contribute to bone loss in the elderly [26]. Calcium absorption is also lower in patients with osteoporosis compared with age-matched controls [27], although the mechanism of this is not clear. Elderly patients with fractures of the neck of the femur have a much higher prevalence of dementia [28] than the normal population although the reasons for this observation are not clear. It may be that demented patients fall more frequently and land heavily but it is also conceivable that they are more likely to sustain a fracture following a fall because of an association with osteoporosis. The above observations suggest a number of questions which the present study sets out to test: (a) Do patients with Alzheimer-type dementia (ATD) have any evidence of abnormal calcium homoeostasis (e.g. low plasma calcium) and, if so, does this correlate with either their clinical or nutritional status? (b) Do patients with A T D have any evidence of altered calcium absorption from the gut? This is important in view of the diminished calcium flux across other peripheral membranes (see above) and, if abnormal, would strengthen the link between ATD and osteoporosis. (c) If either plasma calcium and/or gut calcium absorption is abnormal in ATD, do these changes reflect endocrine changes in ATD, i.e. are these changes secondary to changes in PTH metabolism, vitamin D metabolism (as many patients with ATD are immobile and have reduced exposure to sunlight) or cortisol metabolism (as ATD patients may be hypercortisolaemic and this may affect calcium homoeostasis)? (d) Aluminium has been implicated in the pathogenesis of Alzheimer's disease [29]. Calcium deficiency in rats has been shown to result in increased gastro-intestinal absorption of aluminium and an increased concentration of aluminium in the CNS [30]. We have investigated whether there was any link between indices of calcium homoeostasis and plasma aluminium in A T D patients. (e) Do any changes in calcium homoeostasis
I. N. FERRIER ET AL.
37O
Table. Details of subjects in each of the three groups
Group Sex ratio Age (range) years Weight (kg) MTS (median) score (range) Plasma or serum: Calcium (mmol/1) Albumin (g/1) Corrected calcium (mmol/1) Alkaline phosphatase (U/l) PTH (ng/ml) Cortisol (nmol/1) 25OHD (nmol/1) l,25(OH)2D(pg/ml) Aluminium (/ig/1) Radiocalcium absorption
Alzheimer-type dementia
Multi-infarct dementia
26 20F:6M 76 ±7 (65-86) 54 ±12** 9***(0-29) 2.34±0.13* 40 ± 4** 2.36±0.12 99 ±52 0.8±0.5 332 ±98 32±9 (n=15) 22 ±12 ( n = l l ) 3.8+ 1.3 0.29 + 0.17**
11
5F:6M 76 ±9 (64-90) 63 ±16 11*** (1-29) 2.35 ±0.08* 39 ±4** 2.35 ±0.12 100 ±26
0.38 ±0.24*
Controls 24 18F:6M 74 ±5 (66-86) 65 ±12 36 (35-37) 2.45 ±0.1 44 ±2 2.37 ±0.09 82 ±20 0.8 ±0.3 299 ±97 40 ±25 (n = 15) 20 ±8.4 (n=14) 3.6± 1.5 (n = 19) 0.6±0.30
Means ± SD. *p < 0.05, **p < 0.01, ***p < 0.005, compared to controls.
or absorption in ATD have diagnostic specificity? Subjects All patients were assessed by consultant psychogenatricians (A.F.F., I.G.McK.) and classified using DSM III diagnostic criteria [31], the Hachinski Ischaemic Index (HII) and the 37-item Mental Test Score (MTS) [32]. Fifty per cent of the demented patients were in hospital care and 50% were living in the community (mostly at home) attending either as day- or out-patients. The ages and sexes of groups studied are given in the Table and were classified as follows:
Subjects with acute intercurrent illness, epilepsy or recent fracture were excluded. All subjects had normal renal function (creatinine less than 100 ^mol/1). Subjects on the following drugs were excluded: neuroleptics, antidepressants, anticonvulsants, diuretics, hormonal treatments (e.g. steroids, oestrogens, etc.) and calcium or vitamin D supplements Ethical approval of the study was given by the Joint Ethical Committee of Newcastle District Health Authority/University of Newcastle, patients' relatives gave consent and patients who appeared in any way unco-operative were not studied.
Methods Primary degenerative dementia of Alzheimer type—After an overnight fast, blood (50 ml) was taken from senile onset: All patients scored less than 30 on the a forearm vein via a butterfly cannula. A tourniquet MTS (range 0-29) with HI I scores of < 7. was used for venous access and then removed prior to Multi-infarct dementia: All patients scored less sampling. Subjects then took an oral dose of radioacthan 30 on MTS (range 1-29) with HII scores > 7. tive calcium (185 kBq of 4SCa) washed down by 25 ml and a further venepuncture was carried out Controls: Controls were recruited via the North of water 45 East Alzheimer Disease Society and included if there at 1 h. Ca was measured in pre-dose and 1-h was no history of major psychiatric and/or neurolo- samples on a Packard 2000CA liquid scintillation gical illness. Three subjects had major depressive counter, and radiocalcium absorption calculated episodes recorded some years previously. All sub- using a formula which takes weight into account, as jects were living at home and scored either 36 or 37 on described by Francis et al. [33]. Measurement of calcium absorption using this method correlates well the MTS.
REDUCED ABSORPTION OF CALCIUM IN DEMENTIA
between them. The routine biochemistry (creatinine, electrolytes, calcium, phosphate, a'kaline phosphatase, liver function tests) were in the normal range for all subjects apart from five Calcium, phosphate, albumin and alkaline phospha- ATD and one M I D patients who had minitase were measured in the biochemistry department mally elevated alkaline phosphatase. Both calat Newcastle General Hospital. Corrected calcium cium and albumin were slightly but highly was calculated according to the formula: corrected significantly lower in both groups of demented calcium = measured calcium —[(measured albu- patients compared with controls but corrected m i n - 4 0 ) x0.02]. calcium was almost identical in the three PTH was measured by radio-immunoassay groups. employing a C-terminal directed antiserum. The Radiocalcium absorption was significantly limit of sensitivity of the assay was 100 pg/ml and the reduced in both A T D patients (p
Summary Several reports have suggested that the neurodegenerative change in Alzheimer-type dementia (ATD) may be related to alterations in calcium homoeostasis. The absorption of radiocalcium (4SCa) in 26 ATD subjects and 11 patients with multi-infarct dementia (MID) was compared to 24 normal age- and sexmatched controls. The absorption of radiocalcium was significantly lower in both ATD and M I D when compared to controls. The reduced 4SCa absorption in ATD occurred in the presence of normal plasma concentrations of PTH and vitamin D metabolites and the serum concentrations of calcium and aluminium were in the normal range. The data suggest that the reduced uptake of radioactive calcium observed in A T D is a non-specific derangement.
Introduction Alterations of cellular calcium homoeostasis may be involved in the chain of events that leads to neurodegenerative changes in Alzheimer's disease and recent evidence suggests that such alterations may be expressed in extraneural tissue. The evidence for this statement is diverse and complex and a summary is given below: 1. Impairment of neuronal calcium homoeostasis or of calcium-dependent processes has been implicated in degenerative disorders of the mammalian nervous system such as those associated with ageing or environmental factors [1-3] including toxin-induced nerve-cell death [4]. An elevation of cytoplasmic free calcium in neurones seems to play an important part in brain damage following ischaemic or hypoglycaemic episodes [5, 6]. In neurones at rest the intracellular concentration of free-calcium ions is over four orders of magnitude lower than
extracellular free calcium. This remarkable transmembrane gradient seems essential for maintaining cell integrity because either a decrease in extracellular calcium or an increase in intracellular calcium can lead to neuronal degeneration [7, 8]. Since the surface membrane of neurones is permeable to calcium ions, the low intracellular calcium is achieved by an efficient buffering system for free-calcium ions following their entry into the cytoplasm [9]. The elements of the intracellular calciumbuffering system include the mitochondria, the endoplasmic reticulum, calcium-binding proteins in the cytosol and ATP-linked phosphorylating processes. There is evidence that at least some of the elements of the calcium-buffering systems are affected in Alzheimer's disease. The endoplasmic reticulum associated enzyme aneutral-glucosidase is markedly reduced in Alzheimer-type dementia (ATD) in temporal cortex and hippocampus [10]. Neurones containing parvalbumin (a cytosolic calcium-bindAge and Ageing 1990,19:368-375
REDUCED ABSORPTION OF CALCIUM IN DEMENTIA
ing protein) are markedly reduced in number in ATD cortex [11]. Pyruvate dehydrogenase enzyme complex (PHDC) activity is reduced in ATD and this highly regulated enzyme appears to have an important role in neuronal calcium homoeostasis [12-14]. It is not clear at present if these changes in calcium buffering mechanisms in ATD are primary or secondary events. Aberrant phosphorylation of microtubuleassociated protein is likely to be an important mechanism in the pathogenesis of neurofibrillary tangles [15] and may be related to impaired intracellular calcium homoeostasis with consequent activation of calcium-dependent protein kinases [7]. 2. There is evidence for abnormal intracerebral accumulation of calcium in some dementing disorders. Garruto et al. [1] demonstrated increased calcium deposition in tangle-bearing neurones in the parkinsonism-dementia complex of Guam. Accumulation of aluminium within these neurones also occurs in this disorder and it has been suggested that these changes are due to secondary hyperparathyroidism resulting from an environment rich in aluminium but low in calcium and magnesium. There is evidence for increased calcium in cortex in pathologically confirmed A T D [16]. Mann [17] has shown that the prevalence and severity of calcification in the basal ganglia is greater (for age) in Down's syndrome and in patients with Alzheimer's disease under 75 years of age. However, calcification of the basal ganglia is frequently encountered in patients with a deficiency of parathormone and in such patients epilepsy, parkinsonism and dementia are often present [18]. 3. There is some evidence that alterations in cellular calcium homoeostasis may be expressed in extraneural tissue. Reduced calcium uptake has been demonstrated by cultured skin fibroblasts [19] and by mitogen-stimulated lymphocytes [20] taken from patients with Alzheimertype disease. It is noteworthy that Down's syndrome patients, most of whom develop Alzheimer-type neuropathological changes over the age of 35 [21] have reduced plasma calcium levels [22, 23] and reduced calcium uptake into platelets [24]. Senile plaques and neurofibrillary tangles also occur in normal elderly individuals. Cal-
369
cium absorption from the bowel declines with advancing age [25] and this may contribute to bone loss in the elderly [26]. Calcium absorption is also lower in patients with osteoporosis compared with age-matched controls [27], although the mechanism of this is not clear. Elderly patients with fractures of the neck of the femur have a much higher prevalence of dementia [28] than the normal population although the reasons for this observation are not clear. It may be that demented patients fall more frequently and land heavily but it is also conceivable that they are more likely to sustain a fracture following a fall because of an association with osteoporosis. The above observations suggest a number of questions which the present study sets out to test: (a) Do patients with Alzheimer-type dementia (ATD) have any evidence of abnormal calcium homoeostasis (e.g. low plasma calcium) and, if so, does this correlate with either their clinical or nutritional status? (b) Do patients with A T D have any evidence of altered calcium absorption from the gut? This is important in view of the diminished calcium flux across other peripheral membranes (see above) and, if abnormal, would strengthen the link between ATD and osteoporosis. (c) If either plasma calcium and/or gut calcium absorption is abnormal in ATD, do these changes reflect endocrine changes in ATD, i.e. are these changes secondary to changes in PTH metabolism, vitamin D metabolism (as many patients with ATD are immobile and have reduced exposure to sunlight) or cortisol metabolism (as ATD patients may be hypercortisolaemic and this may affect calcium homoeostasis)? (d) Aluminium has been implicated in the pathogenesis of Alzheimer's disease [29]. Calcium deficiency in rats has been shown to result in increased gastro-intestinal absorption of aluminium and an increased concentration of aluminium in the CNS [30]. We have investigated whether there was any link between indices of calcium homoeostasis and plasma aluminium in A T D patients. (e) Do any changes in calcium homoeostasis
I. N. FERRIER ET AL.
37O
Table. Details of subjects in each of the three groups
Group Sex ratio Age (range) years Weight (kg) MTS (median) score (range) Plasma or serum: Calcium (mmol/1) Albumin (g/1) Corrected calcium (mmol/1) Alkaline phosphatase (U/l) PTH (ng/ml) Cortisol (nmol/1) 25OHD (nmol/1) l,25(OH)2D(pg/ml) Aluminium (/ig/1) Radiocalcium absorption
Alzheimer-type dementia
Multi-infarct dementia
26 20F:6M 76 ±7 (65-86) 54 ±12** 9***(0-29) 2.34±0.13* 40 ± 4** 2.36±0.12 99 ±52 0.8±0.5 332 ±98 32±9 (n=15) 22 ±12 ( n = l l ) 3.8+ 1.3 0.29 + 0.17**
11
5F:6M 76 ±9 (64-90) 63 ±16 11*** (1-29) 2.35 ±0.08* 39 ±4** 2.35 ±0.12 100 ±26
0.38 ±0.24*
Controls 24 18F:6M 74 ±5 (66-86) 65 ±12 36 (35-37) 2.45 ±0.1 44 ±2 2.37 ±0.09 82 ±20 0.8 ±0.3 299 ±97 40 ±25 (n = 15) 20 ±8.4 (n=14) 3.6± 1.5 (n = 19) 0.6±0.30
Means ± SD. *p < 0.05, **p < 0.01, ***p < 0.005, compared to controls.
or absorption in ATD have diagnostic specificity? Subjects All patients were assessed by consultant psychogenatricians (A.F.F., I.G.McK.) and classified using DSM III diagnostic criteria [31], the Hachinski Ischaemic Index (HII) and the 37-item Mental Test Score (MTS) [32]. Fifty per cent of the demented patients were in hospital care and 50% were living in the community (mostly at home) attending either as day- or out-patients. The ages and sexes of groups studied are given in the Table and were classified as follows:
Subjects with acute intercurrent illness, epilepsy or recent fracture were excluded. All subjects had normal renal function (creatinine less than 100 ^mol/1). Subjects on the following drugs were excluded: neuroleptics, antidepressants, anticonvulsants, diuretics, hormonal treatments (e.g. steroids, oestrogens, etc.) and calcium or vitamin D supplements Ethical approval of the study was given by the Joint Ethical Committee of Newcastle District Health Authority/University of Newcastle, patients' relatives gave consent and patients who appeared in any way unco-operative were not studied.
Methods Primary degenerative dementia of Alzheimer type—After an overnight fast, blood (50 ml) was taken from senile onset: All patients scored less than 30 on the a forearm vein via a butterfly cannula. A tourniquet MTS (range 0-29) with HI I scores of < 7. was used for venous access and then removed prior to Multi-infarct dementia: All patients scored less sampling. Subjects then took an oral dose of radioacthan 30 on MTS (range 1-29) with HII scores > 7. tive calcium (185 kBq of 4SCa) washed down by 25 ml and a further venepuncture was carried out Controls: Controls were recruited via the North of water 45 East Alzheimer Disease Society and included if there at 1 h. Ca was measured in pre-dose and 1-h was no history of major psychiatric and/or neurolo- samples on a Packard 2000CA liquid scintillation gical illness. Three subjects had major depressive counter, and radiocalcium absorption calculated episodes recorded some years previously. All sub- using a formula which takes weight into account, as jects were living at home and scored either 36 or 37 on described by Francis et al. [33]. Measurement of calcium absorption using this method correlates well the MTS.
REDUCED ABSORPTION OF CALCIUM IN DEMENTIA
between them. The routine biochemistry (creatinine, electrolytes, calcium, phosphate, a'kaline phosphatase, liver function tests) were in the normal range for all subjects apart from five Calcium, phosphate, albumin and alkaline phospha- ATD and one M I D patients who had minitase were measured in the biochemistry department mally elevated alkaline phosphatase. Both calat Newcastle General Hospital. Corrected calcium cium and albumin were slightly but highly was calculated according to the formula: corrected significantly lower in both groups of demented calcium = measured calcium —[(measured albu- patients compared with controls but corrected m i n - 4 0 ) x0.02]. calcium was almost identical in the three PTH was measured by radio-immunoassay groups. employing a C-terminal directed antiserum. The Radiocalcium absorption was significantly limit of sensitivity of the assay was 100 pg/ml and the reduced in both A T D patients (p
