Magnesium supplementation and cardiovascular disease
Part one of questioning convictions on magnesium
Recently I’ve come to question my convictions about certain supplements and their usefulness in preventing chronic diseases or improving healthspan. While many supplements seemingly offer some benefits, there are often associated downsides that range from financial to physiological. In Framework for pharmacological interventions and supplements, I discuss heuristics for decision-making around exogenous molecules and what considerations can elevate our confidence and expectations about benefits.
Decisions about whether or not to use exogenous molecules come first and foremost down to objectives. Increasingly concisely and precisely being able to define expected outcomes and benefits in the context of specific goals is a crucial prerequisite for the risk-to-benefit calculation.
When speaking of the risk-to-benefit ratio of a drug or a supplement we’re necessarily faced with the requirement of defining these terms. As I’ve eluded to in earlier entries, risk and benefits are mostly thought of in their direct sense, i.e. what “positive” and “negative” effects should we expect with a certain intervention? What often gets overlooked is the contra-positive and -negative of these considerations, i.e. what “positive” and “negative” effects should we expect by not approving of a certain intervention?
Our aspirations should be to have objective markers from which we can infer the efficacy of an intervention. Our level of confidence in efficaciousness should at least partially be predicated on the objective changes to biomarkers or functional measures that serve as proxies for the desired outcomes. By this reasoning, we’re unavoidably faced with the responsibility of justifying why we consider specific measures as proxies for specific objectives. For this necessity and sufficiency are key concepts to internalize, as the conditional and implications relationship between bio- or functional markers and their impact on desired outcomes is essential for accurate assessment of cohort-based relative performance, the efficacy of interventions, and positive progression in relation to baseline and objective.
For a long time, I’ve strongly held to the conviction that magnesium supplementation is beneficial for overall health. Karin Elgar describes in her review Magnesium: A Review of Clinical Use and Efficacy the numerous roles that magnesium plays in maintaining normal physiological functions:
Magnesium is an essential co-factor for various enzymes involved in glycolysis and the Krebs cycle, and mitochondrial magnesium also appears to play an important part in regulating mitochondrial function. As such, magnesium is essential for cellular energy production.
Magnesium acts as an antagonist to calcium, for example in neuromuscular and cardiac function, and helps maintain ionic gradients, i.e. keeping intracellular sodium and calcium low and potassium high. In the nervous system, magnesium is also involved in the regulation of various cell membrane receptors, including N-methyl-D-aspartate (NMDA) and ƴ-aminobutyric acid (GABA) receptors.
Magnesium is involved in DNA, RNA and protein synthesis, and is an essential component in DNA and RNA structure. Magnesium also serves as a structural component in bones and teeth.
Magnesium is known to work as a co-factor in over 300 enzymes and due to its central role in maintaining many bodily functions, we can assume deficiencies or insufficiencies resulting in suboptimal physiological functioning. In fact, as Elgar goes on to describe, while magnesium deficiency is mainly seen in people with severe diseases, insufficiency might not be all that uncommon:
Whilst overt magnesium deficiency is rare, magnesium insufficiency appears to be common and may have significant implications for long-term health. Longer-term complications of magnesium deficiency that commonly go unrecognised include the following: altered glucose metabolism; metabolic syndrome; hypertension; atherosclerosis; osteoporosis; asthma; migraines; pre-eclampsia; cardiovascular disease.
The prevalence of magnesium deficiency seems to vary and is of course dependent on the somewhat questionable cut-offs for deficiency set either by health institutions based on population averages or by authors of studies themselves. Elgar cites in her review a paper estimating the prevalence to be 2.5%-15% and describes a large cohort study from Germany that found the prevalence of hypomagnesemia to be 14.5% and 2% with cut-offs of 0.76 mmol/l and 0.7 mmol/l, respectively. The reference range for plasma magnesium in Finnish adults is set at 0.71 - 0.94 mmol/l.1
Now, it’s important to realize that having inadequate intake is not the same as being deficient. Elgar cites data looking at NHANES (US) and NDNS (UK) cohorts that show median intakes of magnesium to be lower than the recommended. Critics might point out that if this doesn’t result in deficiency, isn’t the recommended intake unnecessarily high? Well, they might be, but there might also be compensatory physiological mechanisms that keep plasma magnesium levels in range despite low intake, which at the same time might cause harm in the long run.
The earlier begs the question: is 0.94 mmol/l more desirable than 0.71 mmol/l when it comes to healthspan? We certainly know that clinical deficiencies are bad, but whether moving from the bottom 25% percentile to the top 25% of the reference range makes any difference isn’t obvious. It’s of course impossible to answer a question like that with any certainty, due to the broad definition of healthspan, but we can try to infer insights from studies looking at magnesium and clinical endpoints.
Firstly, a paper by Pickering et. al. published in 2021 in Nutrients looked at the relationship between magnesium intake and cardiovascular events.2 The study was conducted in the Framingham offspring cohort and controlled thus for a bunch of confounders.3 The participants were stratified based on sodium and potassium intake, both of which positively correlated with magnesium intake. Increased magnesium intake was associated with a decreased risk of cardiovascular disease:
Higher magnesium intake (≥ 240 mg/d), regardless of sodium intake, was associated with substantially lower risks of total CVD (HR = 0.72, 95% CI: 0.56, 0.94; HR = 0.61, 95% CI: 0.42, 0.88 among those with higher and lower sodium intakes, respectively).
The problem with epidemiological studies like this is that the results do not allow us to infer causality. Magnesium intake correlated with fruit and vegetable intake, and fiber intake, which makes any direct conclusions hard to draw. Nevertheless, the direction of the association was at least desirable.
Continuing on the lines of cardiovascular disease, Asbaghi et. al. published a review looking at the effect of magnesium supplementation on blood pressure among patients with type 2 diabetes.4 The review found magnesium supplementation to significantly decrease both systolic and diastolic blood pressure5, but the studies included showed high heterogeneity. A further subanalysis of the results found inorganic magnesium to decrease both systolic and diastolic blood pressure6, while organic did not. Treatment time did also matter with interventions longer than 12 weeks reaching significance.
The review by Asbaghi et. al. does not mention baseline or post-intervention changes to plasma magnesium levels. As I mentioned earlier, having tangible biomarkers to look at is of high importance, especially since we have no idea if the participants had inadequate magnesium status before/at the start of the intervention. In this case, plasma magnesium levels are required to tell if the effect on blood pressure is from a normalization of magnesium status, or something else.
In fact, the review by Asbaghi et. al. is not the only one with similar limitations. Dibaba et. al. published a meta-analysis of randomized controlled trials on magnesium supplementation in participants with insulin resistance, prediabetes, or noncommunicable chronic diseases, in which they didn't include pre- and post-intervention measurements of serum magnesium.7 The trials included lasted from 1 to 6 months and used doses corresponding to 200-450mg daily of elemental magnesium. Based on their analysis, Dibaba et. al. found magnesium supplementation to decrease blood pressure significantly:
The magnesium supplementation resulted in a mean SBP reduction of 4.18 mm Hg and a mean DBP reduction of 2.27 mm Hg for the BP at the end of trial compared with baseline.
A closer look at the individual trials included reveals that, in general, magnesium supplementation was more beneficial when pre-treatment blood pressure was higher vs. lower. Moreover, in only one of the eleven trials included, blood pressure moved in opposite directions in the treatment vs. control group. In the trials where blood pressure decreased, the decrease was always more significant for the treatment group, while in the trials where blood pressure increased, the increases were more significant in the control groups. But in fact, some trials found blood pressure to increase with magnesium treatment.
Contrary to Asbaghi et. al. and Dibaba et. al., Zhang et. al., in their meta-analysis of randomized double-blind placebo-controlled trials named Effects of Magnesium Supplementation on Blood Pressure, did report changes to serum magnesium levels in response to supplementation.8 Magnesium supplementation at a median dose of 368 mg/d (range: 238–960 mg/d) for a median duration of 3 months (range: 3 weeks to 6 months) raised serum magnesium levels on average by 0.05 mmol/L (95% CI, 0.03–0.07). The quartile with the lowest baseline (<0.71 mmol/L) experienced the greatest increase (0.15 mmol/L: CI, 0.11-0.18), while higher quartiles (Q2: 0.72-0.82; Q3: 0.83-0.88; Q4: > 0.88) only had increases of 0.02 mmol/L, on average.
Our dose- and time-response analyses of data from 27 trials showed that oral Mg supplementations at a dose of 200 mg/d or with a duration of 1 month was sufficient to significantly raise serum Mg. Higher doses (≥ 300 mg/d) or longer durations (≥ 2 months) were required to achieve maximal levels of serum Mg by Mg supplementation
Zhang et. al. concluded neatly that magnesium supplementation increased serum magnesium and decreased both systolic and diastolic blood pressure in a dose-dependent fashion:
In this meta-analysis of 34 randomized double-blind placebo-controlled trials involving a total of 2028 participants, we found that oral Mg supplementation led to a significant reduction in both systolic and diastolic BPs (2.00 and 1.78 mm Hg, respectively), although systolic BP and diastolic BP responses differed slightly in dose- and duration-dependent manners, respectively. The BP-lowering effects of Mg supplementation were accompanied by elevated serum Mg levels. Greater reduction in both systolic and diastolic BPs also tended to be present in trials with high quality or low dropout rate. Taken together, our findings support a causal antihypertensive effect of Mg supplementation in adults.
Again, a closer look at the reported numbers reveals that the reductions seemed to coalesce in the individuals with low serum magnesium at baseline. I compiled the numbers from the table provided in the paper:
The increase in serum magnesium is overwhelmingly concentrated to the lowest quartile, and so is the lowering of both systolic and diastolic blood pressure. This will statistically show up as “increased serum magnesium decreases blood pressure in a dose-dependent matter”, but leaves out the fact that all the changes occurred in those who had low magnesium status at baseline, and virtually non of the changes happened in those who had adequate serum magnesium according to the reference ranges.
Examine.com, a quite well-known site that publishes easy-to-read literature reviews of supplements and health topics, gives magnesium an A-grade score for magnesium “notably decreasing” blood pressure.9 They quite rightly point out what seems to be some pre-requisites for observable benefits:
There appears to be a significant reduction in blood pressure assuming one of two conditions is met, either the subject is low in magnesium levels in the body (deficient) or if the subject has elevated blood pressure (140/90 or above), with the latter not requiring a deficiency to precede the blood pressure reducing effects
This, of course, in combination with the other studies, should make us question whether normotensive people with normal serum/plasma magnesium status have anything to gain from supplemental magnesium. In general, the therapeutic window of magnesium is quite wide, i.e. magnesium doesn’t seem to exert any harmful effects, even in individuals with adequate levels. Loose stools are usually the only adverse effect of magnesium supplementation and are easily avoidable by progressively building up to the therapeutic dose and by spreading the dose out. This allows us to be a bit more liberal in our interpretation of results, as small, marginal benefits might be worth pursuing in case there’s little to no downside.
Lastly, I want to address a double-blinded, placebo-controlled, randomized trial that looked at magnesium supplementation and metabolic parameters, and blood pressure. Lee et. al.10 recruited normomagnesemic nondiabetic overweight Korean adults with an average age being around 40 and measured some of their metabolic parameters: fasting glucose, fasting insulin, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, blood pressure, and serum and hair magnesium.
75 of the 155 participants were randomized to receive 300mg of elemental magnesium daily in the form of magnesium oxide for 12 weeks. What’s interesting is that both serum- and hair-magnesium levels decreased in both groups compared to baseline. There were no meaningful changes to either lipid or metabolic parameters, but blood pressure did decrease in both the treatment and control groups:
The decrease in systolic BP was greater in the subgroup of subjects in the intervention group (n=8, -17.1 +- 4.2 mmHg) with systolic BP 140 mmHg at the start of the study than those subjects in the control group (n=16, -6.7 +- 10.7 mmHg); in comparison, those subjects whose initial systolic BP reading was low at baseline did not show a change in systolic BP. Moreover, the decrease in diastolic BP with magnesium supplementation was significantly more than with placebo in both the baseline diastolic BP 80-90 mmHg and >90 mmHg subgroups (-3.4 +- 8.3 vs. -0.8 +- 6.6 mmHg, and -3.4 +- 8.3 vs. -0.8 +- 6.6 mmHg, respectively).
In subgroups where diastolic blood pressure was less than 80mmHg, DBP actually increased on treatment compared to the placebo, although not significantly.
The results by Lee et. al. do demonstrate that decreases in blood pressure with magnesium supplementation are not necessarily dependent on changes to serum magnesium. Blood pressure is an easy marker to measure, and magnesium supplementation seems to offer some benefits in people with hypertension. Due to its relative safety, magnesium as an add-on to other interventions for managing blood pressure might yield extra benefits. Alternatively, in cases of mildly elevated blood pressure, magnesium supplementation and rigorous tracking of blood pressure might be worth considering.
In general, though, the data on magnesium supplementation and its effects on blood pressure and metabolic parameters would suggest benefits if either one of two prerequisites are met: magnesium deficiency/insufficiency and/or hypertension. In these cases, supplementation for cardiovascular health might be a good idea, but otherwise, there doesn’t seem to be a lot of data supporting supplementation. This being said, due to the wide therapeutic window of magnesium and really non-existent adverse effects, taking a bit of a leap of fate and supplementing might yield benefits we’re not equipped to quantify, or non-cardiovascular benefits, something I’ll explore in another entry.
I’m not a physician (yet), and nothing in this post should be interpreted as medical advice and is for general information purposes only. You should not delay seeking medical advice for any problem that you might have.
Take care,
Daniel
A wide range of CVD risk factors were routinely assessed in Framingham. Factors explored as potential confounders in these analyses included age, sex, height, education level, BMI, physical activity, cigarette smoking, alcohol intake, a wide range of individual dietary factors, a Dietary Approaches to Stop Hypertension (DASH) eating pattern score, prevalent hypertension and use of anti-hypertensive medications, as well as time-varying development of hypertension or use of anti-hypertensive medications.
SBP: − 5.78 mmHg, 95% CI: − 11.37 to − 0.19; DBP: − 2.50 mmHg, 95% CI: − 4.58 to − 0.41
SBP: − 8.08 mmHg, 95% CI: − 13.29 to − 2.87; DBP: − 3.77 mmHg, 95% CI: − 5.96 to − 1.58