select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
Research, published in Annals of Internal Medicine, reports that few nutritional supplements or dietary interventions offer protection against cardiovascular disease or death, and that some may cause harm.
Victoria Taylor, Senior Dietician at the British Heart Foundation said:
“This is a review of randomised controlled trials and meta-analyses of randomised controlled trials (RCTs) looking at the effects of nutritional supplements or dietary interventions on death or events such as a heart attack or stroke.
“It can be difficult to get a clear answer about a direct relationship between dietary interventions and death or events. Relying on RCTs alone may not give us the full picture for a dietary intervention like the Mediterranean diet. Unlike drug trials, studies on dietary approaches are very difficult to conduct and may vary widely in their approaches and definitions of the interventions. It would also be all but impossible to carry out a research trial where you carefully controlled the diets of thousands of people over many years.
“In the UK, our dietary guidelines are based on independent review of the full evidence base. This encompasses a range of types of study, not just RCTs, in order to offer the best approach to eating to prevent ill health.”
Prof Susan Jebb, Professor of Diet and Population Health, University of Oxford, said:
“This review confirms the vast majority of previous research that has failed to find benefits of most nutritional supplements. But the suggestion that dietary interventions have no benefit does not reflect the totality of the evidence.
The methodology
“This is an ambitious “review of reviews” which brings together published meta-analyses of trials. It includes a large number of studies and participants but because of the very broad scope of the review, all they could largely do was report the meta-analyses of others. This is not new research.
“Two important issues need to be borne in mind in considering the results. First, any review is only as good as the quality of the underpinning primary research and the authors note the “sub-optimal” quality of the evidence. Second, with so many published analyses about diet and cardiovascular disease, there is a great deal of scope for the inclusion criteria for any review to end up with a biased outcome. Different reviews reach different conclusions. For example, in March this year a review of the Mediterranean Diet concluded “The available evidence is large, strong, and consistent. Better conformity with the traditional MedDiet is associated with better cardiovascular health outcomes, including clinically meaningful reductions in rates of coronary heart disease, ischemic stroke, and total cardiovascular disease” (Martinez-Gonzales et al. Circulation Research. 2019;124:779–798 https://www.ahajournals.org/doi/10.1161/CIRCRESAHA.118.313348). This new, more broad brush review, just a few months later, concludes the Mediterranean Diet does not reduce the risk of premature death of cardiovascular disease. Neither of these two papers report new research, they each provide their own interpretation of what has already been published.
Food supplements
“The latest review includes studies of food supplements and food based dietary interventions. Studies of the former are easier to conduct, usually include a placebo and are reasonably analogous to pharmaceutical trials. The evidence here is consistent with previous reviews and finds no good evidence that vitamin and mineral supplements are associated with a reduction in premature death or in CVD. Except to prevent or correct specific deficiencies (e.g. Vitamin D), or in specific circumstances such as folic acid supplements in early pregnancy to prevent neural tube defects, there is generally good agreement that dietary supplements should not be recommended to the general population.
Dietary interventions
“Randomised controlled trials based on food are much harder to conduct. In a trial it is rare for the intervention group to make the dietary changes recommended to the full extent, while the ‘control’ group, may try to make the changes too, especially if it relates to an issue where there is existing public health recommendations. The net effect is that the difference between groups is nearly always less than the investigators intended. Dietary change is hard; many people struggle to maintain the changes over time and at the end of a trial many people revert to their previous dietary habits. The impact of diet accumulates over many years but because of the costs of running these studies the period of follow up is often short, limiting the potential to detect changes in health and particularly to examine effects on death rates.
“Trials are an important part of the evidence base, but in developing dietary recommendations most expert committees consider the totality of the evidence. This includes observational studies which examine the relationship between habitual diet and health over decades and experimental studies which look at how certain dietary components might affect health to see if there are plausible mechanisms to explain any health effects. This approach provides fairly consistent evidence that diets lower in saturated fat and free sugars and containing more fruits, vegetables and fibre-rich carbohydrates are associated with less heart disease and better health overall.
Why we shouldn’t throw out dietary advice
“This review makes no mention of the importance of body weight. The outcome of dietary interventions to reduce weight is clear and a recent systematic review shows a significant reduction in all-cause mortality, primarily through reductions in cardiovascular disease (Ma et al. BMJ 2017 Nov 14;359:j4849. doi: 10.1136/bmj.j4849 https://www.ncbi.nlm.nih.gov/pubmed/29138133).
“As this new review shows, there is ongoing debate about the best dietary approach to reduce cardiovascular disease, but losing weight means eating fewer calories, which usually involves eating less fat, including saturated fat and less sugar. Many interventions also encourage people to eat more vegetables. There is remarkable consistency in this dietary advice to lose weight with the evidence of a healthy diet from observational studies.
“So despite the limitations about the ideal diet composition to specifically reduce the risk of heart disease, we can be confident that for people who are overweight, dietary change to lose weight will reduce their risk of a premature death. It would be unwise to abandon dietary recommendations that can help people to achieve this. The totality of scientific evidence suggests these dietary changes may also have independent benefits on heart disease, but as the uncertainty in this review highlights, more high quality trials are needed to provide greater confidence.
“So, this review will likely add more heat to the debate about diet and health but sadly no new light.”
Catherine Collins RD FBDA, NHS Dietitian, said:
“The authors of this ambitious ‘umbrella review’ have attempted to pull together definitive evidence on the role of sixteen nutritional supplements and eight dietary approaches in relation to cardiovascular disease, cardiovascular and all-cause mortality. Adding recent randomised controlled trials (RCTs) to the portfolio of evidence already critiqued in previous systematic reviews, the authors conclude that neither the Mediterranean diet, nor some of its key components (such as modifying the amount or proportions of different dietary fats) are beneficial in reducing the risk of cardiovascular or all-cause mortality. Nutritional supplements also failed to show benefit in reducing all-cause or cardiovascular mortality – but perhaps reassuringly neither did they show an increased risk from use, either.
“A potential role for omega-3 fats in reducing the risk of MI and coronary heart disease, and a protective effect of folic acid supplements in reducing risk of stroke by 20% were noted. However, data used to assess the effect of folic acid supplementation was influenced by a large study from China, where dietary intake of folic acid and folates were potentially low and supplements offered greater therapeutic benefit.
“The most striking dietary effect identified by the authors was for a lower salt intake reducing both cardiovascular and ‘all cause’ mortality. Mortality risk was reduced by a third in hypertensives and by 10% in those with normal blood pressure, despite the source of their data – a 2014 Cochrane review by Adler and colleagues1 – demonstrating no significant effect of salt reduction on cardiovascular or all-cause mortality. The authors explain the discrepancy between Adler’s null findings and their positive association using the same data as being due to different statistical methodology to calculate risk. This process inevitably confuses rather than clarifies dietary research.
“The Mediterranean diet and variations (including the Nordic diet, Blue Zones diet) form the basis of international healthy eating guidelines. Using the 2014 data from Liyanage et al2, the authors concluded no reduction in cardiovascular or all-cause mortality from following a Mediterranean diet. These findings conflict with recent systematic reviews3,4. The significant reduction in stroke risk from a Mediterranean style diet identified by the Liyanage group (RR 0.66, 0.48-0.92) – greater than that of folic acid supplementation reported – was omitted from their published findings.
“What should the public make of these findings? First, an umbrella review may well be considered the most comprehensive review of a medical subject, but it cannot address the nitty-gritty of the individual research studies it includes. So whilst this umbrella review confirms what is already known – that nutritional supplements don’t appear to offer much in the way of cardiac health benefit beyond our typical diet – it can’t define the nuances of each study, choosing only to investigate a broad theme – in this case cardiovascular and all-cause mortality.
“Given our longevity as humans, using fixed duration RCTs to define long term benefit (or not) of an intervention may be irrelevant. Our longevity and health influenced by genetic, environmental, diet and lifestyle factors over decades, not typically the weeks or years of a research study. For this reason, cohort studies or epidemiological data may well be more useful to examine nutritional risk at population level. Beneficial non-dietary factors such as lifestyle and exercise cannot be accounted for in this review, yet have significant influence on cardiovascular morbidity and mortality5.
“Professionally, I feel that using an umbrella review to assess dietary research on heart health is like trying to paint a masterpiece using a fence brush – the broad strokes create an impression, but the nutritional nuance, the clinically relevant fine detail, is lost.
“Summary reviews like this discount caveats that influence findings. The flaws of estimating dietary intake using food frequency questionnaires are well established. Few supplement trials expect full compliance with supplement use, accepting that erratic usage confounds ‘intention to treat’ analyses. Supplements may be poorly tolerated. Iron supplements often cause GI symptoms which many people find unacceptable. Many report chewable calcium supplements as unpleasant and so fail to take the recommended dose. Nutrient absorption can be variable (for example, taking an identical dose of vitamin D with the main meal instead of on an empty stomach doubles blood levels; taking vitamin C with iron supplements enhances iron uptake). Some may substitute another supplement which they find more acceptable. Others will take supplement combinations that may create health issues rather than prevent them. None of this variation can be captured by an umbrella review.
“In summary, this review is useful in that a significant number of the population take supplements. Multivitamin and mineral (MVM) supplements are typically viewed as ‘health insurance’ by the general public, yet this review proved usage did not reduce cardiovascular or all-cause mortality risk.
“Perhaps this review reassures us that despite public concern about diet, we appear to derive sufficient cardiovascular benefit from our current diet that can’t be further enhanced by supplementation. However, MVM supplements provide a wide spectrum of nutrients, and we know from the UK NDNS data that some people are borderline deficient in some of these, such as iodine, selenium and vitamin D. As such they may provide health benefits indirectly related to cardiovascular health that is beyond the scope of this review.”
1 Adler AJ et al. Reduced dietary salt for the prevention of cardiovascular disease. Cochrane Database of Systematic Reviews 2014, Issue 12. CD009217
2 Liyanage T et al. Effects of the Mediterranean Diet on Cardiovascular Outcomes-A Systematic Review and Meta-Analysis. PLoS One. 2016: 10; 11: e0159252
3 Eleftheriou D et al. Mediterranean diet and its components in relation to all-cause mortality: meta-analysis. Br J Nutr. 2018; 120: 1081-1097
4 Soltani S et al. Adherence to the Mediterranean Diet in Relation to All-Cause Mortality: A Systematic Review and Dose-Response Meta-Analysis of Prospective Cohort Studies. Adv Nutr. 2019. pii: nmz041. doi: 10.1093/advances/nmz041. [Epub ahead of print] PubMed PMID: 31111871
5 Astell-Burt T et al. Communicating the benefits of population health interventions: The health effects can be on par with those of medication. SSM Popul Health. 2018; 6: 54-62
* ‘Effects of nutritional supplements and dietary interventions on cardiovascular outcomes: An umbrella review and evidence map’ by Safi U. Khan et al. was published in Annals of Internal Medicine at 22:00 UK time on Monday 8 July 2019.
Declared interests
Prof Susan Jebb: “My salary is paid by the University of Oxford with current research funding from the National Institute of Health Research, Wellcome Trust and the British Heart Foundation. I have previously conducted trials of weight management interventions where the research has been funded by WeightWatchers and the Cambridge Weight Plan.”
Catherine Collins: “No conflicts of interest.”
None others received.