Ageing constitutes a risk factor for magnesium deficit. Primary magnesium deficit originates from two aetiological mechanisms: deficiency and depletion. Primary magnesium deficiency is due to insufficient magnesium intake. Dietary amounts of magnesium amounts of magnesium are marginal in the whole population whatever the age. Nutritional deficiencies are more pronounced in institutionalized than in free-living ageing groups. Primary magnesium depletion is due to dysregulation of factors controlling magnesium status: intestinal magnesium hypoabsorption, reduced magnesium bone uptake and mobilization, sometimes urinary leakage, hyperadrenoglucocorticism by decreased adaptability to stress, insulin resistance and adrenergic hyporeceptivity. Secondary magnesium deficit in ageing largely results from various pathologies and treatments common to elderly persons, i.e., non-insulin dependent diabetes mellitus and use of hypermagnesuric diuretics.
Magnesium deficit deficit may participate in the clinical pattern of ageing, particularly in neuromuscular, cardiovascular and renal symptomatologies. The consequences of hyperadrenoglucocorticism- the simplest marker of which is non-response to the dexamethasone suppression test - may include immunosuppression, muscle atrophy, centralization of fat mass, osteoporosis, hyperglycaemia, hyperlipidaemia, atherosclerosis, and disturbances of mood and mental performance through accelerated hippocampal ageing particularly. It seems very important to point out that magnesium deficit and stress aggravate each other in a true 'pathogenic vicious circle', particularly in the stressful state of ageing. The importance of magnesium deficit in the aetiologies of insulin resistance, and the adrenergic, osseous, oncogenic, immune and oxidant disturbances of ageing is still uncertain. Oral physiological magnesium supplementation (5 mg Mg/kg/d) is the best diagnostic tool for establishing the importance of magnesium deficiency. Too few open and double blind studies on the effects of the treatment of magnesium deficiency and of magnesium depletion in geriatric populations have been done. Further study is necessary to assess the true place of magnesium deficit in the pathophysiology of ageing.
In developed countries, magnesium intake is marginal throughout the entire population whatever the age: around 4 mg/kg/day instead of the 6 mg/kg/day recommended to maintain satisfactory balance. The high prevalence of the marginal magnesium deficiency in 15-20 per cent of the population seems consistent with the estimation of nutrient deficiency using probability analysis. These data are particularly relevant to the health of aged persons. However, the elderly population is extremely heterogeneous: diseases, handicaps, physical or psychological impairments expose individuals to more severe nutritional deficiencies. Thus marginal magnesium deficiency is observed in elderly people as well as in the general population, and in free living ageing groups as well as in institutionalized elderly patients, although more pronounced in the latter, whatever countries are considered, America, Australia or Europe. A positive correlation between energy intake and magnesium intake is always observed.
This clinical observation of a decreased adaptability to stress due to ageing relies now on a rich and well-defined animal experimental background. The age-related alterations in brain function particularly concern the hippocampal pyramidal neurones. This part of the limbic system exerts an inhibitory influence on the activity of the hypothalamo-pituitary-adrenal axis. Hippocampal ageing induces a state of hyperglucocorticism. Target cells for glucocorticoids are more highly concentrated in the hippocampus than in any other brain region. Excess corticoid receptor activation mediates neuronal degeneration through an increased influx of calcium into the cells induced by a deleterious increased release of excitatory amino acids - such as kainic acid - associated with a decrease of protective inhibitory amino acids - such as glycine, GABA and taurine. This new hippocampal injury could in turn provoke a new imbalance of the hypothalamo-pituitary-adrenal axis with a 'glucocorticoid cascade' inducing a state of hyperadrenoglucocorticism. The hippocampus is therefore a prime target area for investigation of the events which accompany stress and in particular for the regulation of stress-induced corticosteroid secretion. But the hippocampus is also a basic structure for social life, being involved in mood regulation, control of internal inhibition, memory and learning. Long term potentiation of synaptic transmission in the hippocampus appears as its privileged investigation, too. The differences between normal physiological ageing processes and pathological brain ageing processes may result from ageing-associated susceptibility factors: genetic predispositions, infections agents, environmental toxins or nutritional disorders. Magnesium deficit could be one of these ageing-associated susceptibility factors, particularly through:  the vicious circle initiated between magnesium and stress;  the relation between magnesium and neuroplasticity; and  the links between magnesium and the hippocampus. These links have been observed both in vitro and at pharmacological doses but only once in vivo on a physiological model. Further experimental research is necessary to evaluate the importance of this hypothesis using, for example, either the model of hippocampal ageing accelerated by chronic stress or the kainic acid model under deficient or high magnesium diet. With this latter experimental model in rats various magnesium salts were used in order to increase the magnesium intake. Their effects were subsequently compared according to their respective anions as had been done previously with the model of androgenic seizures in mice. The best protective effects were obtained with magnesium acetyl taurinate which constitutes a powerful combination of taurine, the most neuroprotective inhibitory amino acid, and of magnesium. This impressive animal experimental background on the alterations in stressor reactions due to ageing show the importance of the clinical markers in the failure of adaptability to stress in elderly patients.
Several clinical observations confirm the frequency of hyperadrenoglucocorticism in ageing. Static investigations of glucocorticoid may seem contradictory. Basal plasma concentrations of glucocorticoids, whether measured as 17-hydrocorticosteroids as in the past, or now as immunoreactive cortisol, have been found to show no change wis of the regression lines did not differ significantly, analysis of covariance was carried out which showedth age, to be increased in older men or to be increased in the overall aged population. The highest cortisol values are observed in the stroke subgroup. A significant positive correlation between age and log basal cortisol levels has been found in the entire population as well as in each group, whether healthy, or with dementia of Alzheimer-type, or with stroke. Since the slope a significant increase in log basal cortisol levels with the age of patients (P < 0.001) whatever the type of pathology. In a study without modification in the basal cortisol levels, a pronounced sex difference existed in urinary cortisol excretion, men having higher value. An indirect proof of age-related hypothalamic alteration may rely on the disruption of circadian rhythm for plasma cortisol in elderly subjects.
The best proof of the association with ageing between decreased hypothalamo-pituitary sensitivity and negative feedback regulation by glucocorticoid relies on dynamic investigation, and mainly on the dexamethasone suppression test. Older subjects in all diagnostic categories (normal ageing, dementia, depression) have higher post-dexamethasone plasma cortisol levels. A chronic stressful state is characteristic of the ageing process. The increased stress susceptibility is closely related to the ageing process itself and not so much to any particular age-related pathological conditions such as depression or dementia. The central alterations of glucocorticoid receptors in the hippocampus of aged animals may be mirrored in mononuclear leucocyte corticosteroid receptors; in the human, the mean number of type I and type 11 corticosteroid receptors in mononuclear leucocytes was significantly lower in aged subjects. This situation probably represents a concomitant of the normal ageing process. Analysis of the effects of corticotropin releasing hormone (CRH) on the aged hypothalamo-pituitary-adrenal axis compared with those of dexamethasone confirmed a stepwise decrease in corticotropic sensitivity to the negative feedback signal leading to positive glucocorticoid feedback, an enhanced cosecretion of ACTH secretagogues such as vasopressin or a combination of both Lastly, the responses of plasma ACTH, cortisol and dehydroepiandrosterone to CRH in healthy ageing men are compatible with two hypothesis: (1) a diminished sensitivity of ACTH secretion to negative feedback regulation by glucocorticoid in elderly subjects; (2) an ACTH-independent age-related diminution in adrenal androgen secretion, with preserved glucocorticoid secretion. These alterations in the adrenal biosynthesis of steroids favour cortisol production.
J. Durlach, V. Durlach, P. Bac, Y. Rayssiguier, M. Bara, and A. Guiet-Bara