Magnesium glycinate: Khazaie et al., 2026
Answer capsule
A cross-sectional study of 99 adults in Iran found that higher serum iron levels were associated with increased odds of obstructive sleep apnea, while intermediate magnesium levels were associated with lower odds of OSA. Cross-sectional designs cannot establish causation — this study flags a correlation worth investigating in prospective trials but does not tell us whether correcting magnesium levels would improve sleep outcomes.
Key takeaways
- Participants in the second and third magnesium quartiles had 69-74% lower odds of OSA compared to the lowest quartile (Q2: OR=0.31; Q3: OR=0.26).
- Higher serum iron levels were associated with increased OSA odds (borderline significance at p=0.06 for highest vs. lowest quartile).
- The study used polysomnography — the gold standard for diagnosing OSA — not self-reported sleep quality.
- This is a cross-sectional design: blood was drawn at a single point in time, not before or after diagnosis. It cannot tell us if low magnesium causes OSA, if OSA depletes magnesium, or if both share a common upstream cause.
- Sample size was 99 participants — adequate for detecting associations, but not large enough to control for all confounders or establish dose-response curves.
The study
Khazaie and colleagues published this work in the Journal of Preventive Medicine and Public Health in 2026. They recruited 99 adults from a sleep disorders clinic in western Iran: 54 with polysomnography-confirmed OSA and 45 controls without OSA. The authors measured serum concentrations of six trace elements (iron, magnesium, zinc, copper, selenium, and manganese) via blood draw and compared levels between groups. They used logistic regression to calculate odds ratios for OSA across quartiles of each element and applied a generalized weighted quantile sum regression model to assess the combined effect of all six elements as a mixture. The study was cross-sectional — all measurements were taken at one point in time.
How to read this study
What this paper does well: It used polysomnography to diagnose OSA, which is the diagnostic gold standard — apnea-hypopnea index measured objectively during monitored sleep, not self-reported snoring or fatigue. That eliminates the largest source of misclassification bias in sleep studies. The authors also measured serum levels directly via validated lab assays, not dietary intake questionnaires. Dietary recall is notoriously unreliable; serum levels are objective. They controlled for age, sex, and BMI in their regression models — BMI is a major OSA risk factor, so adjusting for it is critical.
What this paper is missing or weak on: The cross-sectional design is the limiting constraint. We have no idea whether low magnesium preceded OSA or whether OSA altered magnesium metabolism. Sleep fragmentation and hypoxia could theoretically deplete magnesium via stress pathways, or OSA and low magnesium could both be downstream consequences of metabolic dysfunction. The sample of 99 is adequate for detecting large associations but too small to stratify by OSA severity or control for medication use, sleep duration, or dietary patterns beyond the single blood draw. The study also did not measure red blood cell magnesium, which is a more accurate marker of intracellular status than serum magnesium. Serum magnesium is tightly regulated and often normal even when tissue stores are depleted.
How I'd weight this paper: I treat this as hypothesis-generating, not practice-changing. It flags magnesium as worth investigating in a prospective trial — ideally an RCT that supplements magnesium in OSA patients and measures apnea-hypopnea index before and after. On its own, this study does not justify recommending magnesium supplementation for OSA. It does align with mechanistic plausibility (magnesium modulates neuromuscular excitability and autonomic tone, both relevant to airway collapse), which increases my interest in seeing intervention data.
What they found
Serum iron levels were significantly higher in participants with OSA (mean ± SD not reported, p=0.04). Participants in the highest iron quartile had three times the odds of OSA compared to the lowest quartile, though the confidence interval was wide and borderline significant (OR=3.00, 95% CI: 0.96-10.00, p=0.06). The test for trend across quartiles was significant (p=0.02), suggesting a dose-response relationship.
For magnesium, participants in the second quartile had 69% lower odds of OSA compared to the first quartile (OR=0.31, 95% CI: 0.09-0.96), and those in the third quartile had 74% lower odds (OR=0.26, 95% CI: 0.07-0.82). Notably, the fourth quartile — the highest magnesium levels — did not show a significant association, suggesting a U-shaped or threshold relationship rather than a linear one.
When the authors analyzed all six trace elements together as a mixture using weighted quantile sum regression, the combined association was not statistically significant (β=-0.22, p=0.59). This suggests the individual element associations may not be additive or that the sample was underpowered to detect a mixture effect.
What it means for the average man
If you have clinically diagnosed OSA and suspect magnesium deficiency — signs include muscle cramps, eyelid twitches, poor sleep quality independent of apnea — checking serum magnesium is reasonable. This study does not tell us that supplementing magnesium will reduce apnea events, but it flags magnesium as a plausible contributor to sleep physiology in OSA patients. The absence of a linear dose-response (the highest quartile didn't show benefit) suggests that more is not necessarily better and that magnesium's role may be permissive rather than corrective.
For men already supplementing magnesium for other reasons — muscle recovery, nervous system support, bowel regularity — this adds one more data point in favor of adequate intake. Magnesium glycinate at 200-400 mg elemental magnesium per day is well-tolerated and aligns with the intermediate serum levels associated with lower OSA odds in this study.
The caveats
Cross-sectional designs cannot establish causation. We do not know if low magnesium contributes to OSA or if OSA disrupts magnesium homeostasis. The study was conducted in a clinical population in Iran, which may limit generalizability to other populations with different dietary patterns or genetic backgrounds. The authors did not report whether participants were taking magnesium supplements, which could confound serum levels. Serum magnesium is also a poor marker of tissue magnesium status — intracellular measurements would be more informative but are rarely done.
The iron finding conflicts with some prior research suggesting iron deficiency may worsen restless legs syndrome, a condition that co-occurs with OSA. The mechanism linking higher iron to OSA is unclear and may reflect inflammation or metabolic dysregulation rather than a direct causal pathway.
Frequently asked questions
Does this study prove magnesium supplementation will improve OSA?
No. This is a cross-sectional study — it shows an association at one point in time, not a cause-and-effect relationship. To establish that magnesium supplementation reduces apnea events, we would need a randomized controlled trial that gives magnesium to OSA patients and measures apnea-hypopnea index before and after. This study suggests that trial is worth conducting, but the evidence is not yet sufficient to recommend magnesium as an OSA treatment.
What does "cross-sectional" mean and why does it matter?
Cross-sectional means all measurements were taken at a single point in time. The researchers drew blood and measured sleep at the same moment. This design can show correlations but cannot determine which came first — the low magnesium or the OSA. Maybe low magnesium contributes to airway collapse. Maybe OSA disrupts magnesium metabolism. Maybe both are caused by a third factor like chronic stress or poor diet. Without measuring magnesium before OSA develops, or supplementing magnesium and tracking OSA severity over time, we cannot distinguish between these possibilities.
Why didn't the highest magnesium quartile show a benefit?
The study found a U-shaped relationship: intermediate magnesium levels were associated with lower OSA odds, but the highest levels were not. This could mean magnesium's role is permissive rather than dose-dependent — correcting a deficiency helps, but pushing levels above normal does not provide additional benefit. It could also reflect chance variation in a small sample, or confounding by factors the authors did not measure (e.g., supplement use, kidney function). More research is needed to clarify the dose-response curve.
Is 99 participants enough to trust this study?
Ninety-nine participants is adequate to detect large associations but not to control for many confounders or detect small effects. The confidence intervals for the magnesium odds ratios were wide (0.09-0.96 for Q2, 0.07-0.82 for Q3), which means the true effect could be anywhere in that range. A larger study would tighten those intervals and give us more confidence in the point estimate. For hypothesis generation, 99 is reasonable. For practice-changing evidence, we would want 300+ or better yet, a meta-analysis of multiple trials.
Sources
- Khazaie H., Nakhaee S., Manoochehri Z., et al. Association Between Serum Trace Elements and Obstructive Sleep Apnea: Multiple-Exposure Models. J Prev Med Public Health. 2026. PubMed.
- Abbasi B., Kimiagar M., Sadeghniiat K., et al. The effect of magnesium supplementation on primary insomnia in elderly: A double-blind placebo-controlled clinical trial. J Res Med Sci. 2012;17(12):1161-1169.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
