I’ll leave this chart of food sources of magnesium up top as a reminder that plant-based diets are usually abundant in magnesium, as long as you are not restricting your calories:

Below is a placebo-controlled, randomized, cross-over study with 12 older participants, from Germany.

Oral Mg(2+) Supplementation Reverses Age-Related Neuroendocrine And Sleep EEG Changes In Humans, Pharmacopsychiatry, July 2002

Abstract

The process of normal aging is accompanied by changes in sleep-related endocrine activity. During aging, an increase in cortisol at its nadir [lowest point] and a decrease in renin and aldosterone concentration occur. In aged subjects, more time is spent awake and slow-wave sleep is reduced: there is a loss of sleep spindles and accordingly a loss of power in the sigma frequency range. Previous studies could show a close association between sleep architecture, especially slow-wave sleep, and activity in the glutamatergic and GABAergic system. Furthermore, recent studies could show that the natural N-methyl-D-aspartate (NMDA) antagonist and GABA(A) agonist Mg(2+) seems to play a key role in the regulation of sleep and endocrine systems such as the HPA system and renin-angiotensin-aldosterone system (RAAS).

Therefore, we examined the effect of Mg(2+) in 12 elderly subjects (age range 60-80 years) on the sleep electroencephalogram (EEG) and nocturnal hormone secretion. A placebo-controlled, randomised cross-over design with two treatment intervals of 20 days duration separated by 2 weeks washout was used. Mg(2+) was administered as effervescent tablets in a creeping dose of 10 mmol and 20 mmol each for 3 days followed by 30 mmol for 14 days. At the end of each interval, a sleep EEG was recorded from 11 p.m. to 7 a.m. after one accommodation night. Blood samples were taken every 30 min between 8 p.m. and 10 p.m. and every 20 min between 10 p.m. and 7 a.m. to estimate ACTH, cortisol, renin and aldosterone plasma concentrations, and every hour for arginine-vasopressin (AVP) and angiotensin 11 (ATII) plasma concentrations.

Mg(2+) led to a significant increase in slow wave sleep (16.5 +/- 20.4 min vs. 10.1 +/- 15.4 min, < or =0.05), delta power (47128.7 microV(2) +21417.7 microV(2) vs. 37862.1 microV(2) +/- 23241.7 microV(2), p < or =0.05) and sigma power (1923.0 microV(2) + 1111.3 microV(2) vs. 1541.0 microV(2) + 1134.5 microV(2), p< or =0.05 ). Renin increased (3.7 +/- 2.3 ng/ml x min vs. 2.3 +/- 1.0 ng/ml x min, p < 0.05) during the total night and aldosterone (3.6 +/- 4.7 ng/ml x min vs. 1.1 +/- 0.9 ng/ml x min, p < 0.05) in the second half of the night, whereas cortisol (8.3 +/- 2.4 pg/ml x min vs. 11.8 +/- 3.8 pg/ml x min, p < 0.01) decreased significantly and AVP by trend in the first part of the night. ACTH and ATII were not altered.

Our results suggest that Mg(2+) partially reverses sleep EEG and nocturnal neuroendocrine changes occurring during aging. The similarities of the effect of Mg(2+) and that of the related electrolyte Li+ furthermore supports the possible efficacy of Mg(2+) as a mood stabilizer.

They say that normal aging is accompanied by changes in brain waves and hormones which lead to both less time asleep and less-deep sleep. They found that giving magnesium in supplement form partially reversed both of those changes, leading to better sleep.

The amount of magnesium they gave was 30 mmol/day. I think, using this conversion, that works out to about 720 mg. Another site had it at about 750 mg. Either way, that’s a little over twice the RDA for magnesium. If true, it’s a lot. You’d have to be concerned about diarrhea. Magnesium is going to have a more profound effect on people who are deficient or have low levels to start (that wasn’t shared here) so a little might go a long way in people with poor status.

I thought this was interesting:

The similarities of the effect of Mg(2+) and that of the related electrolyte Li+ furthermore supports the possible efficacy of Mg(2+) as a mood stabilizer.

Magnesium and lithium are similar? In their use as mood stabilizers or anti-depressants? How about that.

I’ve written about lithium before. There’s evidence that lithium improves mental health:
An RDA For Lithium? On The Order Of 1 mg/day?, September 2015
Mark My Words, There Will Be An RDA For Lithium In The Future, January 2017