Obviously that publication didn't go over well with the keto crowd, and they found a lot to pick apart (See 1 and 2). The most relevant critiques: Methodology-

Obviously that publication didn't go over well with the keto crowd, and they found a lot to pick apart (See 1 and 2). The most relevant critiques: Methodology-wise, the prospective study seemed pretty poor * Those people filled out questionnaires on their eating patterns on two occasions, six years apart. Their health was followed up for 25 years, allowing for factors that might alter the results, such as smoking, income and diabetes. (the assumption presumably being that they ate the same for the rest of their lives? IMO this seems to make any results pretty much worthless) * low carb in this study is defined as "low carbohydrate consumption (<40%)" - typically LCHF is understood to be 50-100g, so typically 10-20%, and keto is even lower (5%) * The PURE study was cited as part of the meta-analysis but another analysis in the Lancet actually came to the findings that: "Higher saturated fat intake was associated with lower risk of stroke (quintile 5 vs quintile 1, HR 0·79 [95% CI 0·64–0·98], ptrend=0·0498). Total fat and saturated and unsaturated fats were not significantly associated with risk of myocardial infarction or cardiovascular disease mortality." https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)32252-3/fulltext https://www.thelancet.com/journals/lancet/article/PIIS0140-6... * The "years in life expectancy" is very unusual - usually what's presented is relative risk or hazard ratio and absolute risk. This can be found in the data - from the ARIC study, the HR for high carb was 1.16 and for "low" carb was 1.37 - the annual death rate was 1.63%, so the absolute risk was a difference from 1.38% vs 1.88%. These calculations, and a full analysis of the article is available here: http://www.zoeharcombe.com/2018/08/low-carb-diets-could-shorten-life-really/ http://www.zoeharcombe.com/2018/08/low-carb-diets-could-shor... * I noticed that one of the authors listed is Willett, who invented the Mediterranean diet as we know it, and Sarah Seidelmann, the lead author is an outspoken vegan advocate, so it's worth keeping bias in mind when looking at the results. Sadly, the world of nutritional science (especially the epidemiological arm seems to really be as much about personalities as science (part of Ancel Keys' legacy, I suppose) Actually, since I'm a bit tired of listing out points here are a few other interesting rebuttals that are primarily interesting because they point to all the pitfalls of coming to conclusions from observational studies and meta-analyses (and why nutritional reporting is simply the worst): * https://cluelessdoctors.com/2018/08/17/when-bad-science-can-harm-you/ https://cluelessdoctors.com/2018/08/17/when-bad-science-can-... * https://www.linkedin.com/pulse/low-carbs-mortality-john-schoonbee https://www.linkedin.com/pulse/low-carbs-mortality-john-scho... * https://www.psychologytoday.com/us/blog/diagnosis-diet/201809/latest-low-carb-study-all-politics-no-science https://www.psychologytoday.com/us/blog/diagnosis-diet/20180... * https://anhinternational.org/2018/08/22/scientific-attack-on-low-carb-diets/ https://anhinternational.org/2018/08/22/scientific-attack-on... * https://asianwithoutrice.com/making-low-carb-a-murderer-part-2-of-2-broken-from-the-start/ https://asianwithoutrice.com/making-low-carb-a-murderer-part... 1: https://www.reddit.com/r/ketoscience/comments/98kx1w/all_nitpicks_criticism_refutations_and_discussion/ https://www.reddit.com/r/ketoscience/comments/98kx1w/all_nit... 2: https://www.reddit.com/r/ketoscience/comments/981ync/dietary_carbohydrate_intake_and_mortality_a/ https://www.reddit.com/r/ketoscience/comments/981ync/dietary...