Friend or Foe? Dietary Supplements vs All-Cause Mortality

Spin.  Damage control.  That's what you do when you don't get the results you were expecting/hoping.  In this most recent example, a study was published yesterday in the Archives of Internal Medicine which concluded that regular use of dietary supplements increased mortality in older women.  Say what?

You heard that right!  The authors followed 38,772 women average 61.6 years old at baseline who participated in the Iowa Women's Health Study for 18 years.  First, it should be noted that a large proportion of women were already taking at least one multivitamin/mineral supplement at baseline and that the proportion of women taking supplements increased with time (age).  Second, supplement users were less likely to smoke and more likely to take hormone supplements.  They were also more physically active than their non-supplement counterparts.  So it will be very interesting to read the supplement industry's response.

Yet regardless of how the authors divided up the numbers, those who reported taking multivitamins, vitamin B6, folic acid, iron, magnesium, zinc, and copper were found to have a higher total or all-cause mortality regardless of follow up time interval than those who didn't take these supplements.  Calcium use was associated w/lower mortality, contrary to recent findings associating it w/greater heart attack risk.  D, the vitamin du jour, demonstrated neither harm nor benefit.  

But in reality, should we really be surprised?  Previous studies have actually demonstrated harm from vitamin E, vitamin A, and beta-carotene.  Consumption of B vitamins demonstrated no outcome benefit from lowering homocysteine, and possibly even harm from such.  The larger Multiethnic Cohort Study came to a similar conclusion (no all-cause mortality benefit from multivitamin use) earlier this year, but to much less fanfare after following 182,099 participants for 11 years .

As I acknowledged over the weekend in several Q&A sessions regarding the use of supplements, they do have a role in treating/preventing outright disease due to deficiency: vitamin C for scurvy, niacin for pellagra, thiamine for beriberi, iron for microcytic anemia, vitamin B12 +/- folate for macrocytic anemia, folate for spinal tube defects, etc.  However, as the editorialists noted, very few of us are truly deficient and there appears to be U-shaped curve when it comes to vitamin & mineral serum levels and our mortality.  Who knew that Goldilocks was so prescient?

One final series of thoughts.  Remember that this is/was an observational study such that we can only truly develop hypotheses.  We can't determine cause & effect (one possible spin would be to claim that as these women got older & more ill, they resorted to dietary supplements, but it was too little, too late).  Also, because the study population was predominantly Caucasian postmenopausal women, it might not apply to women of different ages & ethnicities, much less men.  Time will tell.

Follow @alvinblin
Tweet