Table of ContentsHow the scientific method works
A brief dictionary of terms we use here
Ways to get confused by scientific studies
How the Scientific Method Works
The scientific method is by definition based on facts. Observations are made as objectively and precisely as possible. The methods are supposed to be described in as detailed a way as is needed for another person skilled in the area to be able to repeat the same study. Naturally, where human greed and pride are involved, games are sometimes played but when it is important, the person reporting the results has to allow competitors into his or her laboratory to see exactly how things were done. Otherwise, that person loses standing in the scientific community. And, if the same results cannot be obtained by other people, the results are not accepted and the person loses standing. (See the history of cold fusion to see how this works.)
Shouting and insults as methods of persuasion have always been allowed but they provide only color and interest for observers.
Appeals to authority (argumentum ab auctoritate) or to popular consensus (Argumentum ad populum) are similarly of no weight.
Albert Einstein, for example, whose theories of special and general relativity have proved correct (so far) despite almost a century of repeated verification that continues to this day, is said to have remarked that he did not believe that "God plays dice with the universe," meaning that he did not accept the probabilistic nature of quantum mechanics. (He apparently believed that, if the underlying physics were understood better, what appeared probabilistic would turn out to be deterministic, as described by Dr. Stephen Hawking.) However, quantum mechanics, as bizarre as it seems to be, does in fact seem to explain correctly how, for example, modern electronic devices work. No one has found an underlying deterministic mechanism to account for it, and no one would reasonably assert that quantum mechanics could not be correct because a noted authority (Albert Einstein) did not accept it. The same reasoning applies to all aspects of scientific discourse.
A brief dictionary of terms we use hereFrom U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015-2020 Dietary Guidelines for Americans. 8th Edition (DGA), pages 90-93:
- Cholesterol. A natural sterol present in all animal tissues. Free cholesterol is a component of cell membranes and serves as a precursor for steroid hormones (estrogen, testosterone, aldosterone), and for bile acids. Humans are able to synthesize sufficient cholesterol to meet biologic requirements, and there is no evidence for a dietary requirement for cholesterol.
- Blood Cholesterol. Cholesterol that travels in the serum of the blood as distinct particles containing both lipids and proteins (lipoproteins). Also referred to as serum cholesterol. Two kinds of lipoproteins are:
- High-Density Lipoprotein (HDL-cholesterol). Blood cholesterol often called "good" cholesterol; carries cholesterol from tissues to the liver, which removes it from the body.
- Low-Density Lipoprotein (LDL Cholesterol). Blood cholesterol often called "bad" cholesterol; carries cholesterol to arteries and tissues. A high LDL-cholesterol level in the blood leads to a buildup of cholesterol in arteries.
- Dietary Cholesterol. Cholesterol found in foods of animal origin, including meat, seafood, poultry, eggs, and dairy products. Plant foods, such as grains, vegetables, fruits, and oils do not contain dietary cholesterol.
- Poultry. All forms of chicken, turkey, duck, geese, guineas, and game birds (e.g., quail, pheasant).
- Lean Meat & Lean Poultry. Any meat or poultry that contains less than 10 g of fat, 4.5 g or less of saturated fats, and less than 95 mg of cholesterol per 100 g and per labeled serving size, based on USDA definitions for food label use. Examples include 95% lean cooked ground beef, beef top round steak or roast, beef tenderloin, pork top loin chop or roast, pork tenderloin, ham or turkey deli slices, skinless chicken breast, and skinless turkey breast.
- Processed Meat & Processed Poultry. All meat or poultry products preserved by smoking, curing, salting, and/or the addition of chemical preservatives. Processed meats and poultry include all types of meat or poultry sausages (bologna, frankfurters, luncheon meats and loaves, sandwich spreads, viennas, chorizos, kielbasa, pepperoni, salami, and summer sausages), bacon, smoked or cured ham or pork shoulder, corned beef, pastrami, pig’s feet, beef jerky, marinated chicken breasts, and smoked turkey products.
- Cross-over study. A study in which a group of participants is divided in two or more subgroups that are asked to take different diets or treatments. After a period of time, they are asked to resume their normal diets or to stop the treatments. Then, after a further period of time, each subgroup is given the diet or treatment that had been assigned to one of the other subgroups. In this way, each participant acts as his or her own "control" subject because each person winds up being tested with each of the diets or treatments in the study. In this way, differences in response due to individual variations cancel out to the extent possible.
- Lipoproteins. Combinations of a lipid like cholesterol with a carrier protein like LDL ("low density lipoprotein") or HDL ("high density lipoprotein"). The terms "high" and "low" density refer to what is seen when blood is spun in a centrifuge at high speed. The high density particles wind up toward the bottom of the tube and the low density particles wind up toward the top of the tube. Please see the wikipedia articles for more information.
- Surrogate endpoint. A variable (such as a type of cholesterol or thickness of the inner lining of the carotid artery) that is known to vary proportionately with the risk of a certain outcome (such as cardiovascular disease).
Ways to get confused by scientific studies
A major source of inaccuracy is simple bias. No matter how well intentioned, someone who has vested interest (such as making money) from a given viewpoint, is likely to find a way to promote that viewpoint. In this article, we have a bias that unprocessed grass-fed beef is healthy. We are trying to be as objective as possible, but it is certainly possible that we have found a way to present a case in favor of this conclusion that is actually not justified by the data. If you think so, do not hesitate to let us know (especially if you can justify your thoughts based on objective data!))
Wikipedia lists more than 130 different logical fallacies, any one of which can invalidate one or all of the conclusions of an otherwise scientific study. A few common ones are listed below:
Association is not causation. The fact, for example, that people in countries where animal protein is commonly eaten also have a high rate of heart attacks does not mean that the animal protein is responsible for the rate of heart attacks. The fact is merely an association. More data are needed to show that the relationship is due to causation.
In public health, Bradford Hill's criteria are one of several methods that have been proposed to try to show causation in a given phenomenon. In microbiology, Koch's postulates are more powerful and can be implemented because the proposed causative agents can be identified and isolated. The relationship of diet and cardiac risk factors is intermediate because the proposed causative agents can be identified but assessing their effect on the cardiac risk factors is more ambiguous than is the case with infectious agents.
- Absence of evidence for an assertion is not evidence against a given assertion (argument from ignorance, argumentum ad ignorantiam).
- Lumping multiple groups together does not permit inferences about the groups that were lumped together (fallacy of division). For example, studies cannot be used to learn the effects of eating meat if they do not distinguish between (that is, they lump together) foods that include processed meat with foods that include only unprocessed meat.
- Unaccounted confounding factors. This is an important concept because it is much less expensive to do a simple study that does not randomize all important variables, and then "mathematically adjust" the data to correct for the known variables that were not randomized, than to do a fully randomized clinical trial. The problem with the less expensive approach is that not all important variables may be known, and the selected mathematical adjustment may not be the correct one for the given data.
This problem is magnified by "meta-analysis" or "metanalysis", a technique in which many different, unrelated studies that study the same question are lumped together in the hope that their deficiencies will average out and their strengths will emerge triumphant. This technique is much less expensive than doing a fully randomized clinical trial, and in fact is widely used when doing such a trial is impossible to do for various reasons. However, a seminal study published in the prestigious New England Journal of Medicine in 1997 reported that meta-analyses of 12 important clinical questions were wrong about 1/3 of the time as assessed by subsequent large-scale randomized clinical trials of the same questions. Being correct only 2/3 of the time is actually pretty good for a technique in the biomedical sciences, but is not nearly good enough to for the technique to be believed blindly in all cases.
- Reproducibility of experiments in the biomedical sciences is (almost) never 100%. Be aware: randomized controlled trials are enormously expensive and time-consuming but are the only known reliable ways to know the truth. Knowlege is achieved in baby steps. For big issues, many identical or very similar randomized controlled trials are needed to make sure we actually know (as best as possible) the truth.
For example, treatment of hypertension reduces the frequency of adverse cardiovascular outcomes in almost all, but not all studies.
This can be seen, for example, in the excellent review by Liu et al. In Figure 4 of that paper, the effects of blood pressure lowering on fatal and non-fatal recurrent stroke are shown by individual study ("trial"), by the type of antihypertensive medication used (diuretic, renin-angiotension system (RAS) inhibitor, and felodipine (a calcium blocker) in the FEVER trial), and all trials combined. You can interpret the results by first locating the vertical line over the "1.0" in the Odds Ratio section of the Figure. This is the line showing the result that shows no benefit with either treatment (drug or placebo). Next, for each trial or aggregate result, locate the horizontal line with a square, circle, or diamond in it that is at the same level. That line represents the 95% confidence interval. By statistical convention, if a result occurs with a probability of 5% or less, it is considered to be significant and not the result of random variation in whatever is being measured. Therefore, by the same convention, any result that lies within the middle 95% is considered due to random variation and is not considered significantly different from the average result. So, if that line crosses the vertical line, that trial or result showed no significant difference from "no benefit either way".
As you look at this Figure, you will see that several trials showed no significant benefit: one of the four diuretic trials, 5 of 6 RAS inhibitor trials, and the one felodipine trial. The aggregate result was not significantly different from "no benefit either way" for the RAS inhibitors and felodipine. However, the aggregate result for all the medications tried was highly signficant and showed, at worst, a 10% benefit with drug treatment. This pattern is not unusual even with large clinical trials, and is the reason that even randomized clinical trials have to be repeated several times to be certain that whatever results are obtained are most likely real and not the result of random, irrelevant, and not reproducible factors.
- A Greener World. https://agreenerworld.org/
- American Council on Science and Health. "USDA's Report on Pesticides & Antibiotics." https://www.acsh.org/news/2018/12/21/usdas-report-pesticides-antibiotics-13679
- American Diabetes Association. "Standards of Medical Care in Diabetesd 2019" Diabetes Care. January 01 2019; volume 42 issue Supplement 1 http://care.diabetesjournals.org/content/42/Supplement_1
- Amin T and JG Mercer. "Hunger and Satiety Mechanisms and Their Potential Exploitation in the Regulation of Food Intake." Curr Obes Rep. 2016; 5: 106–112. doi: 10.1007/s13679-015-0184-5. PMCID: PMC4796328. PMID: 26762623.
- Appel LJ, Sacks FM, VJ Carey, et al. "Effects of Protein, Monounsaturated Fat, and Carbohydrate Intake on Blood Pressure and Serum Lipids. Results of the OmniHeart Randomized Trial" JAMA, November 16, 2005—Vol 294, No. 19 2455-2464. https://jamanetwork.com/journals/jama/fullarticle/201882
- Baudry J, Assmann KE, Touvier M, et al. "Association of Frequency of Organic Food Consumption With Cancer RiskFindings From the NutriNet-Santé Prospective Cohort Study." JAMA Intern Med. 2018;178(12):1597-1606. doi:10.1001/jamainternmed.2018.4357 https://jamanetwork.com/journals/jamainternalmedicine/article-abstract/2707948
- Bierce A. Devil's Dictionary. from Gutenberg Press at http://www.gutenberg.org/files/972/972-h/972-h.htm from Pangyre.org at http://dd.pangyre.org/w/ % COURAGE trial: PCI vs medical therapy
- William E. Boden, M.D.; et al. (April 2007). "Optimal Medical Therapy with or without PCI for Stable Coronary Disease." The New England Journal of Medicine. 356 (15): 1503–1516. doi:10.1056/NEJMoa070829. PMID 17387127.
- Chiu Y-H, Williams PL, Gillman MW et al. for the EARTH Study Team. "Association Between Pesticide Residue Intake From Consumption of Fruits and Vegetables and Pregnancy Outcomes Among Women Undergoing Infertility Treatment With Assisted Reproductive Technology." JAMA Intern Med. 2018;178(1):17-26. doi:10.1001/jamainternmed.2017.5038 https://jamanetwork.com/journals/jamainternalmedicine/article-abstract/2659557
- Chowdhury R, Warnakula S, Kunutsor S, et al. "Association of dietary, circulating, and supplement fatty acids with coronary risk: a systematic review and meta-analysis." Annals of Internal Medicine. 160 (6)(18 March 2014): 398–406. doi:10.7326/M13-1788. PMID 24723079. https://annals.org/aim/article-abstract/1846638/association-dietary-circulating-supplement-fatty-acids-coronary-risk-systematic-review#tab10Div
- Collins R, Reith C, Emberson J. "Interpretation of the evidence for the efficacy and safety of statin therapy." Lancet 2016; 388: 2532–61. http://dx.doi.org/10.1016/S0140-6736(16)31357-5 https://www.thelancet.com/action/showPdf?pii=S0140-6736%2816%2931357-5
- Consumer Reports. http://greenerchoices.org/labels/
- Cruz-Coke R. "Cardiovascular diseases in Chile." Prev Med. 1985 Jul;14(4):541-7. https://www.ncbi.nlm.nih.gov/pubmed/3877926
- Daley CA, Abbott A, Doyle PS, et al. "A review of fatty acid profiles and antioxidant content in grass-fed and grain-fed beef." Nutrition Journal20109:10. https://doi.org/10.1186/1475-2891-9-10 https://nutritionj.biomedcentral.com/articles/10.1186/1475-2891-9-10
- de Souza RJ, Mente A, Maroleanu A, et al. "Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: systematic review and meta-analysis of observational studies." BMJ (Clinical Research Ed.). 351 (11 August 2015): h3978. doi:10.1136/bmj.h3978. PMC 4532752. PMID 26268692. https://www.bmj.com/content/351/bmj.h3978
- DuBroff R and M de Lorgeril. "Cholesterol confusion and statin controversy." World J Cardiol. 2015 Jul 26; 7(7): 404–409. doi: 10.4330/wjc.v7.i7.404 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4513492/
- Echeverría G, Dussaillant C, McGee N, et al. "Mediterranean Diet beyond the Mediterranean Basin: Chronic Disease Prevention and Treatment." http://dx.doi.org/10.5772/intechopen.68937 https://api.intechopen.com/chapter/pdf-download/55474
- Environmental Working Group. "EWG's 2019 Shopper's Guide to Pesticides in Produce™." https://www.ewg.org/foodnews/summary.php
- Grundy SM, Stone NJ, Bailey AL, et al. "2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines." Circulation. 2018;. DOI: 10.1161/CIR.0000000000000625. https://www.ahajournals.org/doi/10.1161/CIR.0000000000000625
- Hawking, S. http://www.hawking.org.uk/does-god-play-dice.html
- Hooper L, Martin N, Abdelhamid A, et al. "Reduction in saturated fat intake for cardiovascular disease." The Cochrane Database of Systematic Reviews. 6 (6)(10 June 2015): CD011737. doi:10.1002/14651858.CD011737. PMID 26068959. https://www.ncbi.nlm.nih.gov/pubmed/26068959 (version written in simplified form for non-medical personnel. A good way to get used to the lingo before you dive into the full report. https://www.cochrane.org/CD011737/VASC_effect-of-cutting-down-on-the-saturated-fat-we-eat-on-our-risk-of-heart-disease
- Jönsson T, Granfeldt Y, Ahrén B, et al. "Beneficial effects of a Paleolithic diet on cardiovascular risk factors in type 2 diabetes: a randomized cross-over pilot study." Cardiovasc Diabetol. 2009; 8: 35. doi: 10.1186/1475-2840-8-35 PMCID: PMC2724493 PMID: 19604407 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2724493/
- Klonoff J. "The Beneficial Effects of a Paleolithic Diet on Type 2 Diabetes and Other Risk Factors for Cardiovascular Disease." Diabetes Sci Technol. 2009 Nov; 3(6): 1229–1232. doi: 10.1177/193229680900300601 PMCID: PMC2787021 PMID: 20144375 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2787021/
- Lau, R. "The Big Fat Lie: How Government Caused the Obesity Crisis." https://71republic.com/2019/01/22/big-fat-lie-government-obesity/
- Lauren E O’Connor, Douglas Paddon-Jones, Amy J Wright, and Wayne W Campbell. "A Mediterranean-style eating pattern with lean, unprocessed red meat has cardiometabolic benefits for adults who are overweight or obese in a randomized, crossover, controlled feeding trial." Am J Clin Nutr 2018;108:33–40. doi: https://doi.org/10.1093/ajcn/nqy075 https://academic.oup.com/ajcn/article/108/1/33/5036105
- LeLorier J, Grégoire G, Benhaddad A, et al. "Discrepancies between Meta-Analyses and Subsequent Large Randomized, Controlled Trials." N Engl J Med 1997; 337:536-542. DOI: 10.1056/NEJM199708213370806 https://www.nejm.org/doi/full/10.1056/NEJM199708213370806
- Liu L, Wang Z, L Gong, et al. "Blood pressure reduction for the secondary prevention of stroke: a Chinese trial and a systematic review of the literature." Hypertension Research (2009) 32, 1032–1040. https://www.nature.com/articles/hr2009139
- Lopez-Jimenez F. "Grass-fed beef: What are the heart-health benefits?" https://www.mayoclinic.org/diseases-conditions/heart-disease/expert-answers/grass-fed-beef/faq-20058059
- Lorber D. "Importance of cardiovascular disease risk management in patients with type 2 diabetes mellitus" Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2014:7 169–183 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4043722/
- Maki KC, Van Elswyk ME, Alexander DD, et al. "A meta-analysis of randomized controlled trials that compare the lipid effects of beef versus poultry and/or fish consumption." Journal of Clinical Lipidology 6(4):352-61 · July 2012 https://www.sciencedirect.com/science/article/abs/pii/S1933287412000153
- Mercola J. "Saturated Fat Phobia Lacks Scientific Basis. May 06, 2014." https://articles.mercola.com/sites/articles/archive/2014/05/06/saturated-fat-phobia.aspx
- Micha R, Wallace SK, and D Mozaffarian. "Red and Processed Meat Consumption and Risk of Incident Coronary Heart Disease, Stroke, and Diabetes Mellitus A Systematic Review and Meta-Analysis." Circulation. 2010;121:2271-2283. doi: 10.1161/CIRCULATIONAHA.109.924977. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2885952/
- Mitchell N, Catenacci V, Wyatt HR, et al. "Obesity: Overview of an Epidemic" Psychiatr Clin North Am. 2011 December ; 34(4): 717–732. doi:10.1016/j.psc.2011.08.005. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3228640/
- Neil A, Cooper J, Betteridge J, et al. "Reductions in all-cause, cancer, and coronary mortality in statin-treated patients with heterozygous familial hypercholesterolaemia: a prospective registry study." Eur Heart J. 2008;2(21):2625-2633. doi:10.1093/eurheartj/ehn422. https://academic.oup.com/eurheartj/article/29/21/2625/530400
- O'Connor LE, Kim JE, and WW Campbell. "Total red meat intake of ≥0.5 servings/d does not negatively influence cardiovascular disease risk factors: a systemically searched meta-analysis of randomized controlled trials." Am J Clin Nutr 2017;105:57–69. doi.org/10.3945/ajcn.116.142521 https://academic.oup.com/ajcn/article/105/1/57/4633933
- Pett KD, Kahn J, Willett WC, et al. "Ancel Keys and the Seven Countries Study: An Evidence-based Response to Revisionist Histories. WHITE PAPER. Commissioned by The True Health Initiative http://www.truehealthinitiative.org/ With emphasis on primary source material, historical records." http://www.truehealthinitiative.org/wordpress/wp-content/uploads/2017/07/SCS-White-Paper.THI_.8-1-17.pdf
- Scientific Steering Committee on behalf of the Simon Broome Register Group. Risk of fatal coronary heart disease in familial hypercholesterolaemia. BMJ. 1991;303(6807):893-896. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1671226/pdf/bmj00148-0031.pdf
- Scientific Steering Committee on behalf of the Simon Broome Register Group. "Mortality in treated heterozygous familial hypercholesterolaemia: implications for clinical management." Atherosclerosis 1999; 142(1): 105–112. https://www.atherosclerosis-journal.com/article/S0021-9150(98)00200-7/fulltext
- Shepherd J, Cobbe SM, Ford I, et al. for the West of Scotland Coronary Prevention Study Group. "Coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group." The New England Journal of Medicine. 333 (20):1301–7. doi:10.1056/NEJM199511163332001. PMID 7566020. https://www.nejm.org/doi/full/10.1056/NEJM199511163332001
- Smith A and Harris C. "Type 1 Diabetes: Management Strategies." Am Fam Physician. 2018 Aug 1;98(3):154-162. https://www.aafp.org/afp/2018/0801/p154.html
- Stevenson JL, Paton CM, and JA Cooper. "Hunger and satiety responses to high-fat meals after a high-polyunsaturated fat diet: A randomized trial." Nutrition 2017; 41: 14-23. https://doi.org/10.1016/j.nut.2017.03.008 https://www.sciencedirect.com/science/article/pii/S0899900717300588
- U.S. Department of Agriculture. "Antibiotic Residues." https://www.nal.usda.gov/fsrio/research/publications/chemical-contaminants/antibiotic-residues
- U.S. Department of Agriculture. "National Organic Program." https://www.ams.usda.gov/about-ams/programs-offices/national-organic-program
- U.S. Department of Agriculture. "Organic 101: What the USDA Organic Label Means." https://www.usda.gov/media/blog/2012/03/22/organic-101-what-usda-organic-label-means
- U.S. Department of Agriculture. "Pesticide Data Program." https://www.ams.usda.gov/datasets/pdp"
- U.S. Department of Agriculture. "Restrictions on Antibiotic Use for Production Purposes in U.S. Livestock Industries Likely To Have Small Effects on Prices and Quantities. November 24, 2015" https://www.ers.usda.gov/amber-waves/2015/november/restrictions-on-antibiotic-use-for-production-purposes-in-us-livestock-industries-likely-to-have-small-effects-on-prices-and-quantities/
- U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015 – 2020 Dietary Guidelines for Americans. 8th Edition. December 2015. Available at https://health.gov/dietaryguidelines/2015/guidelines/.
- Veldhorst MAB, Westerterp KR, and MS Westerterp-Plantenga. "Gluconeogenesis and protein-induced satiety." British Journal of Nutrition 2012; 107(4): 595-600. https://www.cambridge.org/core/journals/british-journal-of-nutrition/article/gluconeogenesis-and-proteininduced-satiety/0A7D80DB672C22FBDBCB3C7B7717662A
- Wang A, Richhariya A, Gandra SR, et al. "Systematic Review of Low‐Density Lipoprotein Cholesterol Apheresis for the Treatment of Familial Hypercholesterolemia" Journal of the American Heart Association 2016; 5(7). https://doi.org/10.1161/JAHA.116.003294 https://www.ahajournals.org/doi/full/10.1161/JAHA.116.003294
- Whelton PK, Carey RM, Aronow WS, et al. "2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines." Journal of the American College of Cardiology 2018: 71 (19, May 2018). DOI: 10.1016/j.jacc.2017.11.006 http://www.onlinejacc.org/content/71/19/e127
- Williams AR. "What are the True Nutritional Benefits of Grass Fed Beef?" https://grassfedexchange.com/blog/what-are-the-true-nutritional-benefits-of-grass-fed-beef-by-allen-r-williams-ph-d
- Wolk A. "Potential health hazards of eating red meat". J Intern Med 2017; 281: 106–122. doi: 10.1111/joim.12543 https://onlinelibrary.wiley.com/doi/pdf/10.1111/joim.12543
- Zanchetti A, Thomopoulos C, Parati G. "Randomized Controlled Trials of Blood Pressure Lowering in Hypertension. A Critical Reappraisal." Circ Res. 2015;116:1058-1073. DOI: 10.1161/CIRCRESAHA.116.303641 https://www.ahajournals.org/doi/pdf/10.1161/circresaha.116.303641