SPECIAL ARTICLE

The Philosophy of Evidence-Based Principles

and Practice in Nutrition
Bradley C. Johnston, PhD; John L. Seivenpiper, MD, PhD;
Robin W.M. Vernooij, PhD; Russell J. de Souza, RD, ScD;
David J.A. Jenkins, MD, PhD; Dena Zeraatkar, MSc; Dennis M. Bier, MD;
and Gordon H. Guyatt, MD

Abstract

The practice of evidence-based nutrition involves using the best available nutrition evidence, together with
clinical experience, to conscientiously work with patients’ values and preferences to help them prevent
(sometimes), resolve (sometimes), or cope with (often) problems related to their physical, mental, and
social health. This article outlines the 3 fundamental principles of evidence-based practice as applied to
the field of clinical nutrition. First, optimal clinical decision making requires awareness of the best
available evidence, which ideally will come from unbiased systematic summaries of that evidence.
Second, evidence-based nutrition provides guidance on how to decide which evidence is more or
less trustworthydthat is, how certain can we be of our patients’ prognosis, diagnosis, or of our
therapeutic options? Third, evidence alone is never sufficient to make a clinical decision. Decision makers
must always trade off the benefits with the risks, burden, and costs associated with alternative manage-
ment strategies, and, in so doing, consider their patients’ unique predicament, including their values and
preferences.
ª 2019 Mayo Foundation for Medical Education and Research. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/) n Mayo Clin Proc Inn Qual Out 2019;3(2):189-199

NUTRITION AND CLINICAL PRACTICE Additional barriers may include physicians’ From the Department of
Community Health and

N
utrition is thought to play a funda- perceptions regarding the effectiveness of and Epidemiology, Dalhousie
mental role in the prevention, treat- adherence to nutritional advice.5 University, Halifax, NS,
ment, and prognosis of both acute Although nutritional interventions may Canada (B.C.J., R.W.M.V.);
Department of Health
and chronic diseases. The field of nutritional potentially offer safe and cost-effective alterna- Research Methods, Evi-
epidemiology, born from epidemiology and tives to pharmaceutical and surgical interven- dence, and Impact (B.C.J.,
other fields of public health, has over the tions for the prevention and management of D.Z., G.H.G.) and
Department of Medicine
past few decades been the foundation to nutri- chronic health problems such as obesity and (G.H.G.), McMaster Uni-
tion research and has had an important influ- type 2 diabetes,6 clinicians may be misin- versity, Hamilton, ON,
ence on the practice of dietitians and dietary formed about the best available evidence. For Canada; Department of
Nutritional Sciences, Fac-
advice globally.1 Medicine, however, has often instance, unfiltered findings in nutritional ulty of Medicine, Univer-
overlooked the role of nutrition in disease pre- science repeatedly make the news headlines; sity of Toronto, ON,
vention and management for a multitude of however, headlines seem to frequently contra- Canada (J.L.S., R.J.dS.,
D.J.A.J.); Toronto 3D
reasons, including a dearth of adequate nutri- dict one another. A recent example includes a Knowledge Synthesis and
tional education, a lack of monetary compen- randomized clinical trial (RCT) of 148 partici- Clinical Trials Unit, Clinical
sation for nutritional advice, and because pants published in the Annals of Internal Nutrition and Risk Factor
Modification Centre (J.L.S.,
much of the current medical practice revolves Medicine that found that low-carbohydrate R.J.dS., D.J.A.J.), Li Ka Shing
around pharmaceutical and procedure- diets were superior to low-fat diets for weight Knowledge Institute (J.L.S.,
oriented care.2,3 Nutrition is not a major focus loss.7 A second study, a network meta-analysis D.J.A.J.), and Division of
Endocrinology and Meta-
during medical training and is often disre- of 48 RCTs, totaling 7286 participants, was bolism, Department of
garded in medical practice apart from fields published just 1 day later in the Journal of Medicine (D.J.A.J.),
such as diabetes and renal failure, where nutri- the American Medical Association and demon- Affiliations continued at
tional care is a mainstay of treatment.4 strated that current evidence shows very little the end of this article.

Mayo Clin Proc Inn Qual Out n June 2019;3(2):189-199 n https://doi.org/10.1016/j.mayocpiqo.2019.02.005 189
www.mcpiqojournal.org n ª 2019 Mayo Foundation for Medical Education and Research. Published by Elsevier Inc. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 MAYO CLINIC PROCEEDINGS: INNOVATIONS, QUALITY & OUTCOMES

difference in weight loss with low- underlying their clinical care, and work with
carbohydrate vs low-fat diets.8 each patient to ensure that chosen interven-
Additional examples of dramatically tions are in the patient’s best interest. Prac-
differing study results come from the findings ticing EBN requires clinicians to understand
of several studies on dietary supplements for how uncertainty about clinical research evi-
the prevention of major cardiovascular dis- dence intersects with an individual patient’s
ease. Early observational studies suggested predicament and preferences regarding the
that vitamin E supplementation reduced car- balance of nutrition and susceptibility to dis-
diovascular death.9 Furthermore, a relatively eases related to nutrition. Herein, we outline
large RCT of 2002 participants compared how EBN proposes to achieve these goals
vitamin E with placebo and found a statisti- and, in so doing, define the nature of EBN.
cally significant 47% relative risk reduction
in cardiovascular death and nonfatal myocar- THE 3 FUNDAMENTAL PRINCIPLES OF EBN
dial infarction with vitamin E supplementa- Evidence-based nutrition involves 3 funda-
tion.10 However, a subsequent larger RCT of mental principles. First, optimal clinical
9541 participants taking vitamin E vs placebo decision-making requires awareness of the
found no difference in myocardial infarction best available evidence that will ideally come
and death from cardiovascular causes,11 and from systematic summaries of the available ev-
a meta-analysis and meta-regression, including idence. Second, EBN provides guidance to
135,967 patients who participated in 19 decide whether evidence is more or less trust-
RCTs, reported that vitamin E not only does worthydthat is, how certain can we be of our
not reduce mortality12 but may also increase patients’ prognosis, of our therapeutic options,
mortality when given in high doses.13 More or of the properties of diagnostic tests? Third,
recently, omega 3 supplementation has evidence alone is never sufficient to make a
demonstrated discrepant results between clinical decision. Decision makers must always
observational studies and meta-analysis of trade off the benefits with the risks, burden,
RCTs among patients at high risk for major and costs associated with alternative manage-
cardiovascular events,14,15 as has been the ment strategies, as well as individuals’ dietary
case with vitamin D for preventing cancer habits and preferences, and, in so doing,
and cardiovascular disease.16-18 consider their patients’ unique predicament
The practice of evidence-based nutrition and values and preferences.19 Each of these
(EBN) involves using the best available nutri- principles assumes that best evidence is readily
tion evidence, together with clinical experi- available to clinical decision makers.
ence, to help patients prevent (sometimes),
resolve (sometimes), or cope with (often) BEST EVIDENCE SUMMARIES
problems related to their physical, mental, Table 1 summarizes the recommendations
and social health, according to their values from 5 major guidelines of experts regarding
and preferences. The EBN approach necessi- prophylactic use of probiotics for the preven-
tates seeking out and understanding clinical tion of Clostridium difficile infection in hospi-
research evidence regarding the role of nutri- talized and nonhospitalized patients, and
tion in health care problems. For those Figure, using a forest plot, represents the evi-
involved in making health care decisions, dence available at the time the recommenda-
EBN encompasses creating implementation tions were made.20
strategies, often among a team of multidisci- Based on 20 trials and more than 3800 pa-
plinary clinicians using a shared decision- tients, the risk reduction of approximately 64%
making framework grounded in the patient’s seems relatively secure by examination, but
values. some doubt remains, for 2 reasons. First, the
At the core of EBN is a care and respect for number of events in absolute terms, 148, is
patients, for whom it will be a disservice if cli- not large. Second, small studies tend to overes-
nicians provide advice that neglects or misin- timate treatment effects, and most of the
terprets research findings. Effective contributing studies are small. Authors
practitioners of EBN strive for a clear and concluded that their certainty in the estimate
comprehensive understanding of the evidence of effect was moderate; that is, the true effect
n n
190 Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005
www.mcpiqojournal.org
EVIDENCE-BASED NUTRITION

TABLE 1. Clinical Practice Guideline Recommendations Regarding Probiotics for the Prevention of Clostridium difficile Infection
Evidence
assessment Evidence assessment
Guideline Year published Recommendation Strength by authors by reviewersa
American Journal of Gastroenterology 2013 Not recommended Strong Low quality 2
Association of Professionals in Infection 2013 Not mentioned Not assigned Not assessed Not applicable
Control and Epidemiology
European Society for Clinical Microbiology 2014b Unclear Not assigned Not assessed 1, 2
and Infectious Diseases
Health Protection Agency/Department of Health 2008 Not recommended Not assigned Not assessed 1, 2
Society for Healthcare Epidemiology of America 2014 Unclear Not assigned Not assessed 1, 2
a
Level 1 signifies a systematic review of randomized controlled trials, and level 2 signifies a single randomized controlled trial.
b
Updated from the 2009 version without updating prevention strategies; however, a section on probiotics is updated. Evidence assessment is conducted using the Oxford
levels of evidence-based medicine.25

is likely to be close to the estimate of effect, but not receive the benefits of this prophylactic
there is a possibility that it differs substantially. therapy. Is this necessary?
Authors suggested that another 2000 patients Until now, there has been considerable
would need to be randomized before decision disagreement among experts, with many rec-
makers could be more secure in the pooled es- ommending against, or not mentioning, pro-
timate of effect.20 Soon thereafter, a large multi- biotic therapy. Why the expert
center trial was published that randomized an disagreement, the lag behind the evidence,
additional 2941 participants, and found no and the inconsistency between recommenda-
benefit of taking probiotics.21 With 21 trials tions and evidence? Similar to other fields,
of 6759 participants and a large 61% relative to the detriment of patients who have not
risk reduction (95% CI, 46%-71%), and received probiotic therapy since 2013, it
without concern regarding safety, a potential may take a decade for the experts to catch
benefit of probiotic therapy remains possible.22 up with the evidence.26 Some concerns with
Two recently updated systematic reviews, the safety of probiotics have been cited; how-
including an individual patient data meta- ever, this is a rare occurrence and tends to be
analysis, identified 11 new RCTs suggesting among immunocompromised patients.27,28
consistent results with the previous reviews, Following EBN principles that limit reliance
particularly in participants taking 2 or more on biological rationale and place far more
antibiotic drugs and in hospital settings where emphasis on empirical evidence, the experts
the risk of C difficile infection is 5% or should have started recommending probiotic
greater.23,24 therapy in 2013. Such a recommendation
Despite the promising 61% relative risk should be weak or conditional until an even
reduction, a systematic review of practice larger definitive multinational trial is
guidelines indicates that authoritative infec- completed.
tious disease organizations do not recommend A similar but opposite example occurred
administration of probiotics despite the fact with dietary guidelines for the general public,
that probiotics, particularly Lactobacillus GG guidelines that for decades recommended
and Saccharomyces boulardii, have the highest reducing dietary fat to prevent cardiometa-
quality of evidence among available prophy- bolic conditions and heart disease. These
lactic strategies (eg, isolating patients with sus- guideline recommendations were based pri-
pected infections, intensive use of disinfecting marily on observational studies, evidence
agents), and few hospitals and health author- considered to be low quality.29-32 Starting in
ities pay to have probiotics available.25 Ran- the 1970s, the US government began to pro-
domized trials, particularly small trials, mote a low-fat diet. The food industry
continue in which half of the patients will as well as school, food assistance, and military

n n
Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005 191
www.mcpiqojournal.org
 MAYO CLINIC PROCEEDINGS: INNOVATIONS, QUALITY & OUTCOMES

Experimental Control Relative risk

Study, year (reference) group, n group, n Weight, % (95% CI)
Events Total Events Total M–H random

Arvola et al, 1999 (32) 1 61 1 58 1.7 0.95 (0.06-14.85)

Beausoleil et al, 2007 (33) 1 44 7 45 3.0 0.15 (0.02-1.14)
Bravo et al, 2008 (34) 0 41 0 45 - Not estimable
Can et al, 2006 (35) 0 73 2 78 1.4 0.21 (0.01-4.37)
Duman et al, 2005 (36) 0 196 1 180 1.2 0.31 (0.01-7.47)
Gao et al, 2010 (37) 9 171 20 84 23.0 0.22 (0.11-0.46)
Hickson et al, 2007 (38) 0 56 9 53 1.6 0.05 (0.00-0.84)
Kotowska et al, 2005 (39) 3 119 10 127 7.9 0.32 (0.09-1.14)
Lönnermark et al, 2010 (40) 1 80 0 83 1.3 3.11 (0.13-75.26)
McFarland et al, 1995 (41) 3 97 4 96 5.9 0.74 (0.17-3.23)
Miller et al, 2008 (47) 4 95 7 94 8.9 0.57 (0.17-1.87)
Miller et al, 2008 (47)* 2 157 0 159 1.4 5.06 (0.25-104.63)
Plummer et al, 2004 (42) 2 69 5 69 4.9 0.40 (0.08-1.99)
Psaradellis et al, 2010 (48) 1 216 4 221 2.7 0.26 (0.03-2.27)
Rafiq et al, 2007 (49) 5 45 22 55 16.1 0.28 (0.11-0.67)
Ruszczynski et al, 2008 (43) 3 120 7 120 7.2 0.43 (0.11-1.62)
Safdar et al, 2008 (44) 0 23 1 17 1.3 0.25 (0.01-5.79)
Selinger et al, 2011 (50) 0 62 0 62 - Not estimable
Surawicz et al, 1989 (45) 3 116 5 64 6.5 0.33 (0.08-1.34)
Thomas et al, 2001 (46) 2 133 3 134 4.0 0.67 (0.11-3.96)
Total (95% CI) 1974 1844 100.0 0.34 (0.24-0.49)
Total events, n 40 108
Heterogeneity: τ2=0.00; chi-square=12.09; P=.79; I2=0%
Test for overall effect: Z=5.87; P<.001 0.01 0.1 1 10 100
Favors experimental Favors control
group group
Relative risk (95% CI) M–H random

FIGURE. Probiotics for the prevention of Clostridium difficileeassociated diarrhea. M-H, Mantel-Haenszel.

dietary programs, followed these recommen- their own preconceptions and by unrepresen-
dations, systematically replacing fat, particu- tative and often lower-quality evidence. This
larly saturated fat from animal products, with first principle of EBN immediately raises
sugar and starch. Evidence suggests that the another question: “How does one recognize
proportion of fat in the US diet decreased by the best evidence?”
25% while the prevalence of obesity and
type 2 diabetes more than tripled.33 Indeed,
evidence from RCTs has now indicated that GUIDES TO CERTAINTY IN ESTIMATES
diets higher in fats (eg, exceeding 35%), and, Summaries of the best evidence for diagnosis,
in particular, replacing carbohydrates with prognosis, or therapeutic interventions present
healthy fats, reduces the risk of cardiometa- evidence on how to interpret test results, pre-
bolic disease.34,35 After almost 4 decades of dict patients’ likely fate, or understand the
low-fat recommendations, the 2015 US Die- impact of alternative management strategies,
tary Guidelines have now placed no upper respectively. Sometimes, such evidence is
limit on total fat consumption.36 trustworthydwe have high certainty in esti-
Rational clinical decisions require system- mates of test properties, patients’ prognosis,
atic summaries of the best available evidence. or treatment effects. At other times, limitations
Without such summaries, cliniciansdexpert in evidence leave us uncertain. Evidence-based
or otherwisedwill be unduly influenced by principles provide guidance to distinguish
n n
192 Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005
www.mcpiqojournal.org
EVIDENCE-BASED NUTRITION

between these situations and the range of cer- up the hierarchy, and RCTs that measure
tainty between them. this effect make up the next step on this
Historically, the question “What is the best hierarchy.
evidence?” was answered with hierarchies of Although RCTs sit on top of the hierarchy
evidence based on study design.37 The most of evidence for therapy and prevention ques-
prominent is the hierarchy related to evidence tions, their limitations for evaluating nutrition
that supports therapeutic interventions questions have been raised by nutritional epi-
(Table 2). Issues of diagnosis or prognosis demiologists. First, trials often need to eval-
require different hierarchies. For studies of uate high vs low intake of a target nutrient;
the accuracy of diagnostic tests, the top of however, when decreasing one nutrient, such
the hierarchy includes studies that enroll pa- as fat, participants will substitute this nutrient
tients about whom clinicians have diagnostic with another, such as simple carbohydrates,
uncertainty and that undertake a blind com- which itself may have health consequences.
parison between the candidate test and a refer- Thus, it is difficult to isolate the effects of a
ence standard. For prognosis, prospective single nutrient. This is a serious concern for
observational studies that accurately docu- studies of the major macronutrients (protein,
ment exposures and objective outcomes, carbohydrate, and fat) but can be more satis-
ideally with blind outcome assessment, and factorily overcome in single-nutrient supple-
follow up all patients during relevant periods ment studies (eg, vitamin E supplement vs
would sit atop the hierarchy. For example, placebo). A second important limitation of
hospitalists are often interested in predictors RCTs in human nutrition is the lack of adher-
of prognosis, such as malnutrition on mortal- ence among trial participants and high
ity in elderly patients after hospital dropout rates related to the often demanding
discharge.38 nature of the intervention, the long period of
Noting the limitations of human intui- follow-up, or both.42,43 Adherence issues,
tion,39 evidence-based principles place unsys- high dropout rates, and expense are particu-
tematic observations based on a small larly relevant to clinical trials of nutrition
number of case reports of individual clinicians attempting to answer effectiveness questions
lowest on the hierarchy (Table 2). Predictions for important outcomes such as cancer or car-
based on physiologic experiments may be diovascular mortality. To capture outcomes of
right but sometimes disastrously wrong; EBN this nature that have a long preclinical time
places such experiments at the next step up course, clinical trials must follow participants
in the hierarchy.40,41 Observational studies for decades to adequately observe an effect.
that measure the apparent effect on patient- Comparatively, prospective observational
important outcomes constitute the next step studies are not faced with the same adherence,
dropout, and expense limitations, allowing
invesigators to better capture and evaluate out-
comes that often take decades to develop.
TABLE 2. Hierarchy of Evidence for Therapeutic Although observational studies have important
Interventions advantages and roles in identifying issues for
Quality of subsequent study and providing guidance
evidence Study design before the conduct of definitive investigation,
Interventional N-of-1 trials reliance on observational studies rather than
studies Randomized controlled trials RCTs may result in misleading inferences
Nonrandomized controlled trials and recommendations.44-46 The shortcomings
Observational Cohort studies of RCTs to evaluate important nutrition ques-
studies Case-control studies tions are well recognized; however, these lim-
Cross-sectional studies itations should not be used to justify placing
Case series excessive trust in the results of typical observa-
Case reports tional studies given their higher risk of
d Background information, expert bias.15,47
opinion, letter to the editor, animal All of the sources of evidence mentioned
research
thus far involve generalizations from groups
n n
Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005 193
www.mcpiqojournal.org
 MAYO CLINIC PROCEEDINGS: INNOVATIONS, QUALITY & OUTCOMES

of patients to an individual, and all are limited of effect of the body of evidence rather than of
in this regard. There are also studies that individual studies. Table 3 summarizes that
involve single patients, and the same strategies framework, formulated by the Grading of
that minimize bias in conventional therapeutic Recommendation Assessment, Development,
trials that involve multiple patients can guard and Evaluation (GRADE) Working
against misleading results in these studies.48,49 Group.50,51 The GRADE approach involves
In the most rigorous n-of-1 RCT, a patient and rating certainty in estimates of the effects of ex-
clinician are blinded to whether that patient is posures or health care interventions (also
receiving active or placebo treatment. Take, for referred to as quality of evidence) as high,
example, the potential use of probiotics for the moderate, low, or very low. Similar to the pre-
treatment of a single patient with irritable vious hierarchy (Table 2), in the GRADE guid-
bowel syndrome, where a patient alternates ance, systematic reviews of RCTs begin as high
probiotics and placebo during several periods, certainty, and reviews of observational studies
and makes quantitative ratings of troublesome are classified as low certainty. The body of ev-
symptoms during each period. The n-of-1 idence becomes less trustworthy, however, if
RCT continues until both the patient and the the individual studies themselves have major
clinician conclude that the patient is or is problems in design and execution (risk of
not obtaining benefit from the target interven- bias); results are imprecise (low event rates
tion, based on statistical evidence. An n-of-1 resulting in uncertainty in the effect estimates
RCT can provide definitive evidence of treat- or CIs include both important benefit and
ment effectiveness for an individual patient48 harm), inconsistent (heterogeneity), or indi-
and is, thus, at the top of the evidence hierar- rect (eg, the population of interest differs
chy. Unfortunately, n-of-1 RCTs are restricted from the population studied); or we have a
to chronic conditions with treatments that act high suspicion of publication bias. When a
quickly without carryover effects, or carryover body of RCT evidence has several of these lim-
effects that washout quickly, and are subject to itations, the certainty in estimates may be low
considerable logistic challenges. We, therefore, or even very low.
must usually rely on studies of other patients Similarly, if treatment effects are suffi-
to make inferences regarding our patient. ciently large and consistent and there is a
This evidence hierarchy is far from abso- dose-response relationship between the expo-
lute, and a more sophisticated framework sure and the outcome of interest, the GRADE
has emerged for judging certainty in estimates approach allows for moderate or even high

TABLE 3. GRADE Approach

1. Establish the initial level of certainty 2. Consider lowering or raising the level of certainty 3. Final level of certainty rating
Certainty in an estimate
Initial confidence in of effect across those
Study design an estimate of effect Lower if Higher if considerations
Randomized controlled High Risk of bias Large effect 4444
trials High
Observational studies Low Inconsistency Dose response 444O
Moderate
Indirectness All plausible confounding and bias would 44OO
reduce a demonstrated effect or would Low
suggest a spurious effect if no effect
was observed
Imprecision 4OOO
Very low
Publication
bias
GRADE ¼ Grading of Recommendation Assessment, Development, and Evaluation.

n n
194 Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005
www.mcpiqojournal.org
EVIDENCE-BASED NUTRITION

certainty ratings from carefully conducted are only indirectly related to a patient impor-
observational studies. For example, observa- tant outcomedbut there is always at least
tional studies have yielded moderate certainty some evidence.61 This consideration may be
in estimates for the association between body particularly relevant in human nutrition,
mass index and the risk of type 2 diabetes in where clinicians have to typically rely on
women and men.52,53 Observational studies observational studies (starting point of low
have produced high certainty in estimates of quality) as the evidence base for dietary guide-
the association between vitamin and mineral lines.62,63 The problem is complicated further
intakes and deficiency diseases.54-56 by often implausible results from assessments
Although there has been criticism from of single nutrients or foods in isolation, such
nutrition epidemiologists about evidence as saturated fat or red meat, assessments that
from observational studies starting at low cer- fail to fully account for the complex interac-
tainty when using the GRADE approach,57 as tions with the dietary and lifestyle patterns in
discussed previously herein, there are many which these nutrients are consumed.44,61,64,65
examples in the medical and nutrition litera- Now that we have the evidence, whatever it
ture of dramatically different results from may be, we can progress to the third principle
RCTs vs earlier observational findings, of EBN: clinical decision making.
including the limited or adverse impact of an-
tioxidants, omega-3 supplementation, and EVIDENCE IS NEVER ENOUGH TO INFORM
reduced dietary fat intake for cardiovascular CLINICAL DECISION MAKING
disease.9,11,14,15,34,35,45,46 Some researchers in Picture a woman with chronic pain from ter-
the nutrition field have argued that RCTs are minal and untreatable cancer. She has come
not feasible for many diet-related questions to terms with her condition, resolved her af-
and that a modified GRADE approach is fairs, said her goodbyes, and wishes to receive
needed.58 However, as outlined previously only palliative care. She develops impaired
herein, there are examples of moderate to glucose tolerance and is at risk for type 2 dia-
high certainty evidence based on observational betes. Evidence that a diet and lifestyle pro-
studies in the nutrition field where large treat- gram reduces the risk of type 2 diabetes
ment effects or a dose-response relationship warrants moderate certainty.66 This evidence
has been demonstrated.52,53,56,59 Moreover, does not, however, dictate that this patient
the limited number and commonly low qual- should receive an interventional program.
ity of RCTs in some areas of nutrition is not Her valuesdemerging from her comorbidities,
a methodological shortcoming of the GRADE social setting, and beliefsdare such that she
approach but a limitation of the evidence would probably prefer to forgo such a
base.60 The GRADE approach has been restricted diet.
endorsed and adopted by more than 120 in- Now picture a second patient, an 85-year-
ternational organizations and societies world- old man with severe dementia who is mute
wide, covering a variety of clinical and and incontinent, has a small social circle,
public health areas. In support of the original and spends his days in apparent discomfort.
intent of the GRADE Working Group, we He is overweight and takes great pleasure in
believe that it is important to maintain stan- overconsuming sweets and desserts. This
dards for assessing the certainty of evidence man develops severe glucose intolerance. Cli-
across health care fields.60 nicians may well be divided in this situation
The evidence-based approach implies on whether to administer a dietary program.
defining a clear course of action for clinicians Again, evidence of treatment effectiveness
addressing patient problems. They should does not automatically imply that the
seek the highest-quality evidence available to restricted diet should be administered.
guide their clinical decisions. The available ev- Finally, picture a third patient, a healthy
idence may warrant a very low certainty rating 30-year-old mother of 2 children who is 2
(ie, extensive uncertainty)dperhaps because months pregnant and develops gestational dia-
the only evidence available is the unsystematic betes. No clinician would doubt the wisdom
observation of a single clinician or physiologic of recommending an exercise program and a
studies that point to mechanisms of action that diet high in fruits, vegetables, and whole
n n
Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005 195
www.mcpiqojournal.org
 MAYO CLINIC PROCEEDINGS: INNOVATIONS, QUALITY & OUTCOMES

grains and low in refined higheglycemic in- Additional clinical skills come in applying
dex carbohydrates to this patient,67 or alterna- evidence, in this instance on the elimination of
tively a low carbohydrate diet.68 This does not foods containing gluten or on potential thera-
mean, however, that an underlying value judg- pies such as pancreatic enzyme supplementa-
ment has been unnecessary. Rather, the values tion for the treatment of celiac. For example,
among the patient, family, and health care an assessment of the applicability of study
providers are concordant, and the benefits so findings occurs when doctors or dietitians
overwhelm the risk and potential inconve- seek the best available evidence and then rely
nience of treatment that the underlying value on their clinical expertise to define features
judgment is unapparent. that affect the applicability of these results to
By values and preferences, we are referring the individual patient. The clinician must
to the collection of goals, expectations, predis- judge the extent to which differences in treat-
positions, beliefs, and abilities and resources to ment (number of dosing regimens, inconve-
make the changes that individuals have for nience of taking the treatment, possibility of
certain decisions that may influence their out- nonadherence to treatment or lifestyle modifi-
comes.69 The explicit enumeration and cation) or patient characteristics (age, comor-
balancing of benefits and risks that are central bidity, and the patient’s cultural, religious, or
to EBN bring the underlying value judgments personal circumstances) may affect estimates
involved in making management decisions of benefit and risk that come from the pub-
into focus and are typically quickly resolved lished literature.
using a shared decision-making model. We note that some of these skillsdthe
Acknowledging that values play a role in sensitivity to the patient’s unique predicament
every important patient care decision high- and the communication skills necessary for
lights our limited understanding of how to shared decision makingdare in many peoples’
ensure that decisions are consistent with an in- minds often not typically associated with
dividual and, where appropriate, societal evidence-based practice. We, however, believe
values. Developing efficient processes for help- that they are at its core. Understanding the pa-
ing patients, clinicians, and allied health pro- tient’s personal circumstances is of particular
fessionals work together toward optimal importance and requires advanced clinical
decisions consistent with patient values and skills, including listening skills and compas-
preferences (eg, decision aids) remains a fron- sion. For some patients, incorporation of pa-
tier for evidence-based decision making. tient values for major decisions will mean a
full enumeration of the possible benefits, risks,
and inconveniences associated with alternative
CLINICAL SKILLS, HUMANISM, AND EBN management strategies. For some patients and
In summarizing the skills and attributes neces- problems, this discussion should involve the
sary for evidence-based practice, take, for patient’s family and other caregivers.
example, a hypothetical scenario that illus-
trates the necessity of getting the diagnosis ADDITIONAL CHALLENGES FOR EBN
right before implementing EBN therapies, in Busy cliniciansdparticularly those early in
which a clinician develops abdominal discom- their development of the skills needed for
fort, bloating, and diarrhea. He self-diagnoses evidence-based practicedwill often find time
that he may have irritable bowel syndrome limitations as their biggest challenge. This
and finds promising evidence from systematic challenge may arise from having inadequate
reviews of a variety of therapeutic options, access to various evidence-based resources.
including probiotics and soluble fiber.70,71 Fortunately, an array of sophisticated
Soon after, realizing the dangers of self- evidence-based information is now available
diagnosis, he visits his family doctor, an expe- for clinicians with online access (eg, PEN:
rienced clinician, to discuss his remaining un- Practice-based Evidence in Nutrition [http://
certainty. The subsequent investigation reveals www.pennutrition.com]), and the pace of
celiac disease and highlights the uselessness of innovation remains rapid.
the evidence for soluble fiber or probiotics for Access to preprocessed information
a condition the patient did not have. cannot, however, address other skills required
n n
196 Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005
www.mcpiqojournal.org
EVIDENCE-BASED NUTRITION

for efficient evidence-based practice. These Commission, American Peanut Council, Barilla, Unilever,
skills include formulating focused clinical Unico/Primo, Loblaw Companies, Quaker (Pepsico), Kellogg
Canada, and WhiteWave Foods. Dr de Souza has served as
questions, matching prioritized questions to
a consultant to the World Health Organization and Cana-
the most appropriate resources, assessing cer- dian Institutes of Health Research; has received grants
tainty in estimates, and understanding how from Canadian Foundation for Dietetic Research, Canadian
to apply results to clinical decision making. Institutes of Health Research, and Hamilton Health Sciences
Although these skills take time to learn, the Centre/PHRI; and received travel expenses and honorarium
from the World Health Organization, McMaster Children’s
reward in terms of efficient and effective prac-
Hospital, and University of Toronto. Dr Bier has served
tice can more than compensate. on the scientific advisory board of ConAgra has consulted
This paper has dealt primarily with deci- (including travel and expenses) for Ajinomoto, International
sion making at the level of the individual pa- Council on Amino Acid Science, University of Texas, Ox-
tient. Evidence-based approaches can also ford University Press, Ferrero, ILSI, Nutrition Growth Solu-
tions, Watson Green LLC, American Society for Nutrition.
inform health care policy-making, day-to-day
Dr Bier has also received payment for lectures from Nestle,
decisions in public health, and systems-level Indiana University, Purdue University, the International Con-
decisions, such as those facing hospital man- ference on Nutrition and Growth, International Council on
agers. In each of these areas, EBN can support Amino Acid Science, Texas A&M University, Nicaragua As-
the appropriate goal of gaining the greatest sociation of Internal Medicine, ILSI, Nutrition Society of
Australia, Society for Risk Assessment, Lorenzini Foundation,
health benefit from limited resources.
Washington University, CrossFit Foundation, Prolacta
Bioscience, Virginia Society for Parenteral and Enteral Nutri-
Abbreviations and Acronyms: EBN = evidence-based tion, Ferrero, Society for Nutrition, and Pfizer. Dr Johnston
nutrition; GRADE = Grading of Recommendation Assess- is a member of the GRADE Working Group and has
ment, Development, and Evaluation; RCT = randomized received travel expenses from Cornell University and Texas
clinical trial A&M University. Dr Guyatt is a member of the GRADE
Working Group.
Affiliations (Continued from the first page of this
article.): St Michael’s Hospital, Toronto, ON, Canada; Publication dates: Received for publication February 20,
Department of Research, Netherlands Comprehensive 2019; accepted for publication February 27, 2019.
Cancer Organisation, Utrecht (R.W.M.V.); Population
Health Research Institute, Hamilton, ON, Canada (R.J.dS.); Correspondence: Address to Bradley C. Johnston, PhD,
Department of Medicine, Faculty of Medicine, University Department of Community Health and Epidemiology, Fac-
of Toronto, ON, Canada (D.J.A.J.); and Children’s Nutrition ulty of Medicine, Dalhousie University, Halifax, Canada
Research Center, Baylor College of Medicine, Houston, TX (bradj49@gmail.com).
(D.M.B.).

REFERENCES
Potential Competing Interests: Dr Seivenpiper serves on 1. Hu FB, Willett WC. Current and future landscape of
the board of the European Fruit Juice Association Scientific nutritional epidemiologic research. JAMA. 2018;320(20):2073-
Expert Panel; is a consultant to Perkins Coie LLP, Tate & 2074.
2. Adams KM, Lindell KC, Kohlmeier M, Zeisel SH. Status of nutri-
Lyle, and Wirtschaftliche Vereinigung Zucker e.V; receives tion education in medical schools. Am J Clin Nutr. 2006;83(4):
grant support from Canadian Institutes of Health Research, 941S-944S.
Diabetes Canada, PSI Foundation, Banting and Best Diabetes 3. Kris-Etherton PM, Akabas SR, Bales CW, et al. The need to
Centre, Canadian Nutrition Society, American Society for advance nutrition education in the training of health care
Nutrition, INC International Nut and Dried Fruit Council professionals and recommended research to evaluate imple-
Foundation, National Dried Fruit Trade Association, The mentation and effectiveness. Am J Clin Nutr. 2014;99(5 suppl):
Tate and Lyle Nutritional Research Fund at the University 1153S-1166S.
4. Devries S, Dalen JE, Eisenberg DM, et al. A deficiency of
of Toronto, Glycemic Control and Cardiovascular Disease nutrition education in medical training. Am J Med. 2014;
in Type 2 Diabetes Fund at the University of Toronto (a 127(9):804-806.
fund established by the Alberta Pulse Growers), and the 5. McClinchy J, Dickinson A, Barron D, et al. Practitioner and
Nutrition Trialists Fund at the University of Toronto (a patient experiences of giving and receiving healthy eating
fund established by the Calorie Control Council); is on advice. Br J Community Nurs. 2013;18(10):498-504.
the speakers’ bureaus of Diabetes Canada, Mott’s LLP, Dairy 6. Hallberg SJ, McKenzie AL, Williams PT, et al. Effectiveness and
Farmers of Canada, FoodMinds LLC, PepsiCo, The Ginger safety of a novel care model for the management of type 2 dia-
Network LLC, International Sweeteners Association, Nestlé, betes at 1 year: an open-label, non-randomized, controlled
study. Diabetes Ther. 2018;9(2):583-612.
Pulse Canada, Canadian Society for Endocrinology and
7. Bazzano LA, Hu T, Reynolds K, et al. Effects of low-
Metabolism, GI Foundation, Abbott, Biofortis, American So- carbohydrate and low-fat diets: a randomized trial. Ann Intern
ciety for Nutrition, Health Sciences North, and Physicians Med. 2014;161(5):309-318.
Committee for Responsible Medicine; and has received in- 8. Johnston BC, Kanters S, Bandayrel K, et al. Comparison of
kind food donations to support a randomized controlled weight loss among named diet programs in overweight and
trial from the Almond Board of California, California Walnut obese adults: a meta-analysis. JAMA. 2014;312(9):923-933.

n n
Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005 197
www.mcpiqojournal.org
 MAYO CLINIC PROCEEDINGS: INNOVATIONS, QUALITY & OUTCOMES

9. Knekt P, Reunanen A, Järvinen R, et al. Antioxidant vitamin 28. Hempel S, Newberry S, Ruelaz A, et al. Safety of probiotics
intake and coronary mortality in a longitudinal population study. used to reduce risk and prevent or treat disease. Evid Rep Tech-
Am J Epidemiol. 1994;139(12):1180-1189. nol Assess. 2011;200:1-645.
10. Stephens NG, Parsons A, Schofield PM, et al. Randomised 29. Keys A, Menotti A, Aravanis C, et al. The seven countries study:
controlled trial of vitamin E in patients with coronary disease: 2,289 deaths in 15 years. Prev Med. 1984;13(2):141-154.
Cambridge Heart Antioxidant Study (CHAOS). Lancet. 1996; 30. Turpeinen O, Karvonen MJ, Pekkarinen M, Miettinen M,
347(9004):781-786. Elosuo R, Paavilainen E. Dietary prevention of coronary heart
11. Yusuf S, Dagenais G, Pogue J, et al. Vitamin E supplementation disease: the Finnish Mental Hospital Study. Int J Epidemiol.
and cardiovascular events in high-risk patients: the Heart Out- 1979;8(2):99-118.
comes Prevention Evaluation Study Investigators. N Engl J 31. Mozaffarian D, Ludwig DS. The 2015 US dietary guidelines:
Med. 2000;342(3):154-160. lifting the ban on total dietary fat. JAMA. 2015;313(24):2421-
12. Miller ER III, Pastor-Barriuso R, Dalal D, et al. Meta-analysis: 2422.
high-dosage vitamin E supplementation may increase all-cause 32. Mozaffarian D. Dietary and policy priorities for cardiovascular
mortality. Ann Intern Med. 2005;142(1):37-46. disease, diabetes, and obesity. Circulation. 2016;133(2):187-225.
13. Bjelakovic G, Nikolova D, Gluud C. Meta-regression analyses, 33. Ludwig DS. Lowering the bar on the low-fat diet. JAMA. 2016;
meta-analyses, and trial sequential analyses of the effects of sup- 316(20):2087-2088.
plementation with beta-carotene, vitamin A, and vitamin E singly 34. Appel LJ, Sacks FM, Carey VJ, et al; OmniHeart Collaborative
or in different combinations on all-cause mortality: do we have Research Group. Effects of protein, monounsaturated fat, and
evidence for lack of harm? PLoS One. 2013;8(9):e74558. carbohydrate intake on blood pressure and serum lipids: results
14. Poole CD, Halcox JP, Jenkins-Jones S, et al. Omega-3 fatty acids of the OmniHeart randomized trial. JAMA. 2005;294(19):2455-
and mortality outcome in patients with and without type 2 dia- 2464.
betes after myocardial infarction: a retrospective, matched- 35. Estruch R, Ros E, Salas-Salvadó J, et al; PREDIMED Study Inves-
cohort study. Clin Ther. 2013;35(1):40-51. tigators. Primary prevention of cardiovascular disease with a
15. Aung T, Halsey J, Kromhout D, et al. Omega-3 Treatment Tria- Mediterranean diet. N Engl J Med. 2013;368(14):1279-1290.
lists’ Collaboration. Associations of omega-3 fatty acid supple- 36. Dietary Guidelines Advisory Committee. Scientific Report of
ment use with cardiovascular disease risks: meta-analysis of the 2015 Dietary Guidelines Advisory Committee. http://
10 trials involving 77 917 individuals. JAMA Cardiol. 2018;3(3): www.health.gov/dietaryguidelines/2015-scientific-report Pub-
225-234. lished 2015. Accessed February 1, 2019.
16. Yin L, Ordóñez-Mena JM, Chen T, Schöttker B, Arndt V, 37. Murad MH, Asi N, Alsawas M, et al. New evidence pyramid.
Brenner H. Circulating 25-hydroxyvitamin D serum concentra- Evid Based Med. 2016;21(4):125-127.
tion and total cancer incidence and mortality: a systematic re- 38. Buscemi S, Batsis JA, Parrinello G, et al. Nutritional predictors of
view and meta-analysis. Prev Med. 2013;57(6):753-764. mortality after discharge in elderly patients on a medical ward.
17. Zhang R, Li B, Gao X, et al. Serum 25-hydroxyvitamin D and Eur J Clin Invest. 2016;46(7):609-618.
the risk of cardiovascular disease: dose-response meta-analysis 39. Nisbett R, Ross L. Human Inference. Englewood Cliffs, NJ: Pren-
of prospective studies. Am J Clin Nutr. 2017;105(4):810-819. tice-Hall; 1980.
18. Manson JE, Cook NR, Lee IM, et al; VITAL Research Group. 40. Guyatt GH, Sackett DL, Cook DJ. Evidence-Based Medicine
Vitamin D supplements and prevention of cancer and cardio- Working Group. Users’ guides to the medical literature, II:
vascular disease. N Engl J Med. 2019;380(1):33-44. how to use an article about therapy or prevention. A. Are
19. Napodano R. Values in Medical Practice. New York, NY: the results of the study valid? JAMA. 1993;270(21):2598-2601.
Humana Sciences Press; 1986. 41. Mann JI. Evidence-based nutrition: does it differ from evidence-
20. Johnston BC, Ma SS, Goldenberg JZ, et al. Probiotics for the based medicine? Ann Med. 2010;42(7):475-486.
prevention of Clostridium difficile-associated diarrhea: a system- 42. Satija A, Yu E, Willett WC, Hu FB. Understanding nutritional
atic review and meta-analysis. Ann Intern Med. 2012;157(12): epidemiology and its role in policy. Adv Nutr. 2015;6(1):5-18.
878-888. 43. Stampfer M. Observational epidemiology is the preferred
21. Allen SJ, Wareham K, Wang D, et al. Lactobacilli and bifidobac- means of evaluating effects of behavioral and lifestyle modifica-
teria in the prevention of antibiotic-associated diarrhoea and tion. Control Clin Trials. 1997;18(6):494-499; discussion 514-516.
Clostridium difficile diarrhea in older inpatients (PLACIDE): a 44. Ioannidis JP. Implausible results in human nutrition research.
randomised, double-blind, placebo-controlled, multicentre trial. BMJ. 2013;347:f6698.
Lancet. 2013;382(9900):1249-1257. 45. Brignardello-Petersen R, Ioannidis JPA, Tomlinson G, Guyatt G.
22. Daneman N. A probiotic trial: tipping the balance of evidence? Surprising Results of Randomized Trials. Users’ Guides to the Med-
Lancet. 2013;382(9900):1228-1230. ical Literature. 3rd ed. New York, NY: McGraw-Hill; 2015.
23. Johnston BC, Lytvyn L, Lo CK, et al. Microbial preparations 46. Young SS, Karr AF. Deming, data and observational studies: a
(probiotics) for the prevention of Clostridium difficile infection process out of control and needing fixing. Quality Control Appl
in adults and children: an individual patient data meta-analysis Stat. 2013;58(1):31-32.
of 6,851 participants. Infect Control Hosp Epidemiol. 2018; 47. Harris WS, Kennedy KF, Maddox TM, Kutty S, Spertus JA. Mul-
39(7):771-781. tiple differences between patients who initiate fish oil supple-
24. Goldenberg JZ, Yap C, Lytvyn L, et al. Probiotics for the preven- mentation post-myocardial infarction and those who do not:
tion of Clostridium difficile-associated diarrhea in adults and chil- the TRIUMPH Study. Nutr Res. 2016;36(1):65-71.
dren. Cochrane Database Syst Rev. 2017;12:CD006095. 48. Guyatt G, Sackett D, Taylor DW, et al. Determining optimal
25. Lytvyn L, Mertz D, Sadeghirad B, et al. Prevention of Clostridium therapy: randomized trials in individual patients. N Engl J Med.
difficile infection: a systematic survey of clinical practice guide- 1986;314(14):889-892.
lines. Infect Control Hosp Epidemiol. 2016;37(8):901-908. 49. Shamseer L, Sampson M, Bukutu C, et al. CONSORT exten-
26. Antman EM, Lau J, Kupelnick B, et al. A comparison of results of sion for N-of-1 Trials (CENT) 2015: explanation and elabora-
meta-analyses of randomized control trials and recommenda- tion. BMJ. 2015;350:h1793.
tions of clinical experts: treatments for myocardial infarction. 50. Guyatt GH, Oxman AD, Kunz R, et al. What is “quality of ev-
JAMA. 1992;268(2):240-248. idence” and why is it important to clinicians? BMJ. 2008;
27. Meini S, Laureano R, Fani L, et al. Breakthrough Lactobacillus 336(7651):995-998.
rhamnosus GG bacteremia associated with probiotic use in an 51. Balshem H, Helfand M, Schünemann HJ, et al. GRADE guide-
adult patient with severe active ulcerative colitis: case report lines, 3: rating the quality of evidence. J Clin Epidemiol. 2011;
and review of the literature. Infection. 2015;43(6):777-781. 64(4):401-406.

n n
198 Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005
www.mcpiqojournal.org
EVIDENCE-BASED NUTRITION

52. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk 63. World Health Organization. Guideline: sugars intake for adults
of type 2 diabetes mellitus in women. N Engl J Med. 2001; and children. http://www.who.int/nutrition/publications/guidelines/
345(11):790-797. sugars_intake/en Published 2015. Accessed May 3, 2019.
53. Wang Y, Rimm EB, Stampfer MJ, et al. Comparison of abdom- 64. Sievenpiper JL, Dworatzek PD. Food and dietary pattern-based
inal adiposity and overall obesity in predicting risk of type 2 dia- recommendations: an emerging approach to clinical practice
betes among men. Am J Clin Nutr. 2005;81(3):555-563. guidelines for nutrition therapy in diabetes. Can J Diabetes.
54. Institute of Medicine. Dietary Reference Intakes for Calcium, Phos- 2013;37(1):51-57.
phorus, Magnesium, Vitamin D, and Fluoride. Washington, DC: 65. Johnston BC, Alonso-Coello P, Bala MM, et al. Methods for
National Academies Press; 1997. trustworthy nutritional recommendations NutriRECS (Nutri-
55. Institute of Medicine. Dietary Reference Intakes for Thiamin, Ribo- tional Recommendations and accessible Evidence summaries
flavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Composed of Systematic reviews): a protocol. BMC Med Res
Biotin, and Choline. Washington, DC: National Academies Press; Methodol. 2018;18(1):162.
1998. 66. Schellenberg ES, Dryden DM, Vandermeer B, et al. Lifestyle
56. Institute of Medicine. Dietary Reference Intakes for Vitamin A, interventions for patients with and at risk for type 2 diabetes:
Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manga- a systematic review and meta-analysis. Ann Intern Med. 2013;
nese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc. Washing- 159(8):543-551.
ton, DC: National Academies Press; 2001. 67. Viana LV, Gross JL, Azevedo MJ. Dietary intervention in patients
57. Schwingshackl L, Knüppel S, Schwedhelm C, et al. Perspective: with gestational diabetes mellitus: a systematic review and
NutriGrade: a scoring system to assess and judge the meta-analysis of randomized clinical trials on maternal and
meta-evidence of randomized controlled trials and cohort newborn outcomes. Diabetes Care. 2014;37(12):3345-3355.
studies in nutrition research. Adv Nutr. 2016;7(6):994-1004. 68. Huntriss R, Campbell M, Bedwell C. The interpretation and
58. Schwingshackl L, Knüppel S, Schwedhelm C, et al. Reply to JJ effect of a low-carbohydrate diet in the management of
Meerpohl et al. Adv Nutr. 2017;8(5):790-791. type 2 diabetes: a systematic review and meta-analysis of
59. Institute of Medicine. Dietary Reference Intakes for Calcium and randomised controlled trials. Eur J Clin Nutr. 2018;72(3):
Vitamin D. Washington, DC: National Academies Press; 2011. 311-325.
60. Meerpohl JJ, Naude CE, Garner P, Mustafa RA, Schünemann HJ. 69. Montori VM, Brito JP, Murad MH. The optimal practice of
Comment on “Perspective: NutriGrade: A Scoring System to evidence-based medicine: incorporating patient preferences in
Assess and Judge the Meta-Evidence of Randomized practice guidelines. JAMA. 2013;310(23):2503-2504.
Controlled Trials and Cohort Studies in Nutrition Research”. 70. Didari T, Mozaffari S, Nikfar S, et al. Effectiveness of probi-
Adv Nutr. 2017;8(5):789-790. otics in irritable bowel syndrome: updated systematic review
61. Ioannidis JPA. The challenge of reforming nutritional epidemio- with meta-analysis. World J Gastroenterol. 2015;21(10):3072-
logic research. JAMA. 2018;320(10):969-970. 3084.
62. Kushi LH, Doyle C, McCullough M, et al. American Cancer So- 71. Moayyedi P, Quigley EMM, Lacy BE, et al. The effect of fiber sup-
ciety guidelines on nutrition and physical activity for cancer pre- plementation on irritable bowel syndrome: a systematic review
vention. CA Cancer J Clin. 2012;62(1):30-67. and meta-analysis. Am J Gastroenterol. 2014;109(9):1367-1374.

n n
Mayo Clin Proc Inn Qual Out June 2019;3(2):189-199 https://doi.org/10.1016/j.mayocpiqo.2019.02.005 199
www.mcpiqojournal.org