Critics say interpersonal variation not given enough weight in new study
The paper, published yesterday in Annals of Internal Medicine, concluded that most vitamin and mineral supplements on the market, as well as most dietary interventions, lack enough evidence supporting their roles in reducing cardiovascular health risks.
Lead author Safi U. Khan, MD, of West Virginia University, said in a video released by the American College of Physicians to promote the study results that “only a few of the 16 nutritional supplements, and one of the eight dietary modifications evaluated had some proven effects in cardiovascular risk reduction.”
These findings came from a systematic review of 277 published clinical trials involving nearly a million subjects around the world.
Lack of observational studies
Trade groups were quick to slam the study. Steven Mister, president and CEO of the Council for Responsible Nutrition (CRN), a dietary supplement industry trade group, called the paper “a coordinated, all-out assault on nutrition, and the critical role it plays in maintaining health and reducing the risk of chronic disease.”
One of the main limitations of the study, Mister argued, was the lack of observational studies.
“Exclusion of observational, or epidemiological, studies is a major limitation, as epidemiological data are critical and serve as the basis of many recommendations made in the Dietary Guidelines for Americans,” he said.
Commenting independently on the paper, Dr Paul Coates, a board member of the American Society for Nutrition and former director of the Office of Dietary Supplements at the National Institutes of Health, did not see any issue with the exclusion of observational studies.
It’s an oft-raised criticism of systematic reviews of clinical trials, he told us, but “because of the inherent risks of confounding in observational studies, I do not share this view.”
Study ‘dismisses decades of nutrition research’
Trade groups slam review saying most supplements, even dietary interventions, offer no cardiovascular protection.
Clinical trials, which involve study participants going through a standardized study protocol so researchers can see if an intervention may impact any health markers or outcomes, measure different data than do observational studies, which look at trends seen from a sample population with no intervention. In other words, some researchers say that comparing the data and measurements of intervention studies with observational studies is like comparing apples to oranges.
“Rather, where there are discrepancies between the findings from interventional and observational studies, one should look for reasons to understand these differences,” Dr Coates added. “If both kinds of studies were to point generally in the same direction, I’d say they were more likely to be correct.”
The problem of summarizing data form large swaths of people
Dr Coates said that the paper’s results mirror those of many previous studies in pointing out the paucity of reliable evidence that dietary supplements can reduce the risk of cardiovascular disease. “In that respect, there is really nothing new here,” he said.
Echoing his comments was Alice Lichtenstein, D.Sc., a volunteer expert for the American Heart Association (AHA) and Gershoff Professor of Nutrition Science and Policy at Friedman School, Tufts University.
“With regard to nutrient supplements there is a robust history of testing associations identified in observational studies with randomized controlled trials,” she told us.
“In many cases the associations are not confirmed, likely do to confounding with factors that co-vary. For example, people who regularly take nutrient supplements (regardless of the type) tend to eat higher quality diets, engage in higher levels of physical activity and are less likely to use tobacco products.”
The two agree that a major limitation is extracting conclusions from the data of large swaths of people.
“Unfortunately, this study is hampered by variability in quality of the evidence in individual trials, as the study authors point out,” Dr Coates said.
Additionally, “we ought to be looking at whether or not subgroups of the populations under study could profit from using a particular dietary supplement ingredient or dietary intervention,” he added.
Asking the right questions?
The notion that there isn’t evidence on how supplements may slash cardiovascular disease risk is not new. Last year, two papers were published with the same conclusion, though focusing on multivitamins. One was a meta analysis in the Journal of the American College of Cardiology, and the other in Circulation: Cardiovascular Quality and Outcomes.
Jeffrey Blumberg, PhD, a senior scientist in the Antioxidants Research Laboratory at the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University, suggested last year that meta-analyses on supplements and heart health might have asked the wrong questions.
“There are ways of looking at quality of life. You could look at whether the subjects had better physiological function like lower blood pressure or better immune function or cognitive function. Those would be a better questions to ask, rather than looking at hard outcomes like CVD risk or overall mortality,” Blumberg said about last year’s studies that downplayed supplements’ potential in cardiovascular health.
“The flip side of that, for me, is that there may well be subgroups of the same populations who are at risk of harm – one simply cannot tell from the average outcome across a broad swath of individuals.”
‘Testing any diet or supplement in a broad population without acknowledging interindividual variability seems like a recipe for failure’
Together with the paper, the journal also published an editorial titled Dispense with Supplements for Improving Heart Outcomes by doctors Amitabh C. Pandey, MD, and Eric J. Topol, MD, of the Scripps Research Translational Institute in California responding to Dr Khan’s team’s paper.
“Geographic considerations among the studies included in Khan and colleagues' analysis are also notable,” they wrote. The benefits seen in folate supplements, for example, may have come from the inclusion of one study from China, where a folate-rich diet is not routine.
“Differences in geography, dose, and preparation—most studies rely on food diaries, which are based on a person's memory of what they consumed— raise questions about the veracity of the data.”
“Testing any diet or supplement in a broad population without acknowledging interindividual variability seems like a recipe for failure,” they argued. “Especially because most trials are not randomized, are not of sufficient duration, or do not have enough hard outcome events.”
There is the issue of background diet, an example brought up by AHA’s Dr Lichtenstein.
“Response to an intervention, whether it be a nutrient supplement or diet modification, is dependent on baseline intakes,” she said.
“For example, if the habitual intake of omega-3 fatty acid or folate is high supplementation it is less likely to have an effect than if intake is low, were there to be an effect.”
With regard to diets, she added, one can’t automatically expect adding nuts or olive to a Western dietary pattern will have the same effect as adding them to a Mediterranean dietary pattern, “it would have to be tested.”
Considerations for future research
“Perhaps, however, the biggest difference that needs to be considered in the future is the individual,” according to the editorial by Drs Pandey and Topol.
“Only recently with machine learning of large data sets, which include multimodal data on physical activity, sleep, medications, demographic characteristics, intake and timing of all foods and beverages, and gut microbiome constituents, have we begun to learn that the use of any specific diet or supplement is likely to have markedly heterogeneous effects,” they added.
