Discussion
Overall, we found a significant association between low magnesium intake and depression, especially in younger adults. The increased prevalence of depression was confined to the lowest levels of magnesium intake. Nonetheless, the effect is very strong, with a >50% higher rate of depression in the lowest quintile of intake compared with those consuming greater amounts. A very different pattern among seniors was observed (Figure 2 and Table 3). First, the overall rates of depression were lower. Second, the spread in rates across the levels of magnesium intake were much higher for younger adults (23% to 37%) than in seniors (15% to 21%). Third, the lowest quintile of intake among seniors did not have the highest prevalence of depression. Rather, the highest rates occurred in the group with the highest intake. Although the adjusted odds of depression were significantly greater in the group with the highest intake compared with the lowest quintile, there was no clear dose–response relationship, and the clinical significance of this finding is uncertain. The large sample size available for analysis (372 to 435 seniors in each quintile of magnesium intake) may be responsible for making a small or even negligible effect seem statistically significant. Even if we discount the seemingly adverse effect of high magnesium intake in seniors, however, there is little doubt that the increased prevalence of depression with low intakes seen in younger subjects is absent after age 65.
Our data show over half of adults do not consume adequate amounts of magnesium. This finding is similar to other US population studies. Magnesium excretion increases while absorption decreases with age because of various chronic diseases and decreased intake of foods high in magnesium. Compared with imipramine, magnesium supplementation was effective in treating depression in older adults with hypomagnesemia and type 2 diabetes in a randomized controlled trial. The current analysis differs by suggesting a detrimental effect of higher magnesium intake in older adults. The differences may be because this is a study of a general US population with low intake rather than a group selected for low serum magnesium concentrations and diabetes.
Little is known about the mechanism of the possible effect of magnesium on depression. Even less is known about this mechanism in the elderly and why the association might differ between age groups. The current findings might be due to unidentified confounders, reverse causality, or data error. The PHQ was comparable with the Geriatric Depression Screen in a convenience sample of elderly primary care patients. The existence of an emotional paradox in elders, however, in which older adults experience higher levels of well-being despite cognitive and physical decline, may influence how depression is identified and scaled in this group and may make the PHQ less sensitive.
Several studies have looked at whether overall dietary pattern is more important than specific nutrients when considering the influence of nutrition on depression. Among adults with mood disorders, mineral intakes may be associated with psychiatric disorders more so than vitamin intakes. We cannot rule out that a specific dietary pattern or combination of nutrients would show a synergistic effect and a stronger relationship with depression than magnesium alone. For instance, residents of Greece, where most people follow a Mediterranean diet, have a lower rate of depression and mental disorders. Changing dietary patterns takes time, however, as well commitment on the part of the patient. Emphasis on the consumption of foods high in magnesium, such as green leafy vegetables, legumes, nuts, seeds, and whole grains, could offer a dietary approach to controlling symptoms of depression. Advocating for increased magnesium intake through food can lead to a healthier overall diet and might be attractive to patients who have previously experienced unwanted side effects from medications for depression. Magnesium supplementation may be effective in as little as 1 week but may lead to gastrointestinal upset such as nausea, vomiting, or diarrhea in some people. However, toxic concentrations are unlikely to occur when the recommended dose of magnesium is given and kidney function is normal. Whether increased dietary magnesium would lead to an improvement in symptoms as quickly as supplements is unknown.
Strengths and Limitations
This study has several strengths. The sample size was large and representative of almost 180 million American adults. The analysis includes the most recent available data from 2007 to 2010 and captures both dietary and supplemental intake. Although the assessment of magnesium intake at only one time point may not reflect long-term intake, trained interviewers collected the data, and the methods were validated and consistent over the 4 years of data collection. In addition to social, demographic, behavioral, and clinical covariates, this analysis also controlled for folate intake, which has not been included in many previous analyses.
As with all observational studies, we cannot exclude the possibility of additional confounding not accounted for by the analysis. Likewise, we cannot rule out reverse causality to explain the association between magnesium intake and depression. In other words, poor dietary intake of magnesium could be a result of mental illness.