Multivitamins: An Expensive Placebo

According to Johns Hopkins, anywhere from 33% to 50% of adults in the US take a multivitamin or supplement. We spend anywhere from $12 to $50 billion on these products annually. The US Preventive Services Task Force (USPSTF), in reviewing the impact of multivitamins on all-cause mortality, found “insufficient evidence for determine benefits or harms.” Perhaps the new study, reported by JAMA Network Open, involving nearly 400,000 participants with 164,000 deaths, will finally put the multivitamin vampire to death. But I wouldn’t bet on it. 

To study whether daily multivitamin (MV) use was “associated with lower mortality risk among generally healthy US adults,” researchers combined three large studies. 

• The National Institutes of Health–AARP Diet and Health Study (NIH-AARP), begun in 1995 involving questionnaires mailed to AARP members aged 50 to 71
• The Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) a screening trial begun 9n 1993 with participants aged 55 to 74 with a dietary questionnaire assessing MV use
• The Agricultural Health Study (AHS) involving licensed pesticide applicators and their spouses, age 18 or older, from Iowa and North Caroline begun in 1993

Because these studies were not primarily interested in MV use, researchers had to “harmonize” the differences in how the use of MVs was noted, along with all potential confounders we would anticipate. [1] Participants were followed from their baseline assessment until the “date of death, loss to follow-up, or the end of the study period.”

Before jumping to conclusions, we must take a moment to consider the meaning of the term “harmonize.” This means taking the varying ways these studies categorize multivitamin use and confounders and meshing them into the categorizations reported in this mashup. It is a moment when subjectivity and objectivity may carry equal weight – a time when bias, intended or not, is introduced. 

Looking across the socioeconomic and confounder data presented, it is fair to say that the cohort was generally healthy, but with over 90% of participants non-Hispanic whites, it does not truly reflect the US. 

Daily users of MVs were more often female, college-educated, non-smokers, and “more likely to use individual supplements and have lower BMI and better diet quality.” As we would anticipate, MV users lead more “healthful” lifestyles. So it should be surprising that not only was there no “mortality benefit” accruing to these individuals over 20 years but that daily MV use “was associated with 4% higher mortality risk” than for non-users.  

That’s right, if the numbers are to be believed, multivitamins not only drain your wallet but increase your risk of dying. No wonder the FDA states 

“This product is not intended to diagnose, treat, cure, or prevent any disease."

While the researchers point to the strength of large numbers and a very long follow-up period with repetitive assessment of MV use, the study's limitations are significant. It is not generalizable to all of the US population because of the exceedingly limited representation of anyone, not a Non-Hispanic White. The researchers also suggest that the findings are uncertain because the impact of MV on our health may lag for a more significant time period than they examined. Twenty years should be adequate to see something, though. 

Will the words of primary author Erikka Loftfield, PhD, MPH, National Cancer Institute, finally drive a stake into the heart of multivitamin use?

"We did not find evidence to support improved longevity among healthy adults who regularly take multivitamins." 

- Erikka Loftfield, PhD, MPH, National Cancer Institute

Doubtful, extremely doubtful. There are any number of investigators where hope springs eternal. From the ironic Pieter Cohen, MD, of Cambridge Health Alliance

"It's hard for me to be worried about [the mortality signal] given that we have randomized controlled trials ... saying multivitamins don't kill you." 

To the more measured thoughts of the invited commenters suggesting that mortality was not the only significant health impact and that multivitamins might still confer some healthy years of life, what we term health span. But even they suggest that

“Micronutrients come most healthfully from food sources.”

Just like our mothers told us.

[1] Potential confounders included “sex, age, self-reported race and ethnicity, education level, smoking status and intensity, body mass index (BMI)… marital status, physical activity level, alcohol intake, coffee intake, Healthy Eating Index … other individual supplement use, and family history of cancer.” 

Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts JAMA Network Open DOI: 10.1001/jamanetworkopen.2024.18729