Wikipedia:Peer review/Ketogenic diet/archive1

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

Ketogenic diet

(more info)

This peer review discussion has been closed.
The ketogenic diet is a high-fat diet for treating epilepsy. It is currently a Good Article and I'd like to push it on towards FAC. I would like the article to be accessible to and interesting for the "general reader", who may not know much about epilepsy or human metabolism. So opinions from non-medical-experts are most welcome. I'm pretty confident in the quality of the sources and verifiability of the text. How's the balance of the article? Are some sections too long or short? Were some bits hard to follow or boring? Where is the prose not engaging or, worse, gramatically poor?

BTW: I typically review medical articles (currently got 2009 flu pandemic on my TODO list to look at) but if editors here want me to review their article in exchange, let me know and I'll be happy to have a look.

Thanks, Colin°Talk 16:50, 12 October 2009 (UTC)

Comments by Eubulides (talk · contribs)
  • This looks like a very good article. Here are a few comments (this is not a proper review, but I figured I'd write what I can for now).
  • The claim "The ketogenic diet is endorsed by ... US insurance companies." is supported by 3 sources. The first, Lefevre & Aronsen 2000 (PMID 10742367), doesn't contain any wording that I can see that supports the claim: it is a systematic review that was partly supported by Blue Shield, and which is positive about the KD, but that's not the same as Blue Shield endorsing KD. The other two sources are a clinical guideline for one insurance company and a coverage policy manual for another; this is uncomfortably close to original research.
  • I briefly looked for 3rd-party sources to support the claim, and found two statements that to some extent indicate the contrary. Kossoff et al. 2009 (PMID 19535814) and Zupec-Kania & Spellman 2008 (PMID 19033218) both say that medium-chain triglyceride is not reimbursable by insurance companies. The latter source does briefly mention "insurance preauthorization" without saying what's involved; presumably it's not always a slam-dunk.
  • I found the "Interdisciplinary team" section of Zupec-Kania & Spellman 2008 (PMID 19033218) to be quite a useful summary of the procedures (and hassles) of implementing KD, written more for the nutritionist/nurse's point of view rather than the doctor's POV. This is a valuable viewpoint that I suspect is currently underrepresented in the article.
  • While we're on the subject of Zupec-Kania and colleagues, it may be helpful to cite both the work mentioned above, and Zupec-Kania & Zupanc 2008 (PMID 19049580), in preference to the older Zupec-Kania et al. 2004 (ISBN 1588292959) source currently cited. (I haven't read any of these sources; it's just a guess.)
  • And that leads to a different question. In November Epilepsia published a special 133-page issue on KD (vol. 49, issue s8), but none of the articles in that issue are cited here. Is this due to lack of access, or did none of these recent articles actually make the cut, or what?
Hope this helps. As you can probably tell I am not really up to speed here, and am just browsing sources. Eubulides (talk) 07:45, 15 October 2009 (UTC)
Thanks very much. I'll try to find alternative info on the US insurance situation, or reword/drop it. The MCT insurance issue is covered by the article and is specific to the MCT version of the diet. I'll check out the other papers you mention to see if I have them or are making best use of them. I agree that the nutritionist/nurse POV is much less covered by the literature and I'll review whether the article's balance is right here. This therapy is far more staff-intensive than prescribing pills. I'll review the paper against the older book chapter I've used.
Wrt the Epilespia supplement, see Wikipedia talk:Reliable sources (medicine-related articles)/Archive 3#Journal supplement. I'll have another read of that set of articles. IIRC I was disappointed the mini-review-articles in the set didn't really improve on the sources I already had, and the latest-research-findings articles in the set weren't so usable as sources on WP.
Do you think I should update sources just for the sake of citing something newer? For example, PMID 17332207 and PMID 17241206 both date from 2007 but are comprehensive reviews. Where the more recent PMID 18823325 and PMID 19535814 have modified or added information, I've replaced and revised the text using them, but I haven't sought to change the source if nothing has changed.
Colin°Talk 08:28, 15 October 2009 (UTC)
The Zupec-Kania & Zupanc 2008 paper is 3.5 pages compared to 19 pages in the 2004 book. However, it does appear to have some newer stuff, particularly about patients with feeding difficulties. I guess I'm guilty of concentrating so much on the medical aspects that I haven't emphasised enough the importance of the health/diet aspects. I'll have a look at expanding this from both sources. I'm trying to get a copy of Zupec-Kania & Spellman 2008. I have most relevant papers but that one seemed to slip by. Colin°Talk 20:50, 15 October 2009 (UTC)
Wow, I'd forgotten all about that earlier thread. No, of course there's no need to update a citation to a newer one if it is just an abbreviated version of an older, more-authoritative one, and it sounds like you've done exactly the right thing, namely, check the newer/shorter/less-authoritative version to see whether it has a new result or trend that isn't in the older one, and cite it for that. (Clearly I need to read all these sources!) Eubulides (talk) 19:38, 15 October 2009 (UTC)
  • "In the UK, the National Institute for Health and Clinical Excellence states that the diet should not be recommended for adults with epilepsy because there is insufficient evidence of efficacy." The NICE guideline is ambiguous and quite plausibly could be read to be more negative than that. The guideline says "The ketogenic diet should not be recommended for adults with epilepsy. [C]" The "[C]" means that this recommendation is "Based directly on level III evidence or extrapolated from level I or level II evidence". A plausible way to interpret this is that the evidence says that the ketogenic diet should not be recommended (as opposed to the evidence basis being too weak for a recommendation). I suggest removing the text "because there is insufficient evidence of efficacy", because that phrase is not clear from the cited source. Eubulides (talk) 22:51, 16 October 2009 (UTC)
I've removed that clause and added a bit more from elsewhere. Although there was a large study of adults in 1930, there have been only two modern studies (A prospective study in 1999 with 11 patents and a retrospective study in 2003 with 45 patients). The focus of research in older patients seems to be moving towards the higher-carb variants, though the consensus paper requests "larger, randomized better understand the efficacy and tolerability of these diets, how they compare to the classic KD, and potential use in adults." Colin°Talk 15:58, 17 October 2009 (UTC)

I have some qualms about File:Ketogenic diet over 12 months.png. First, its horizontal axis is not uniform, which is a no-no; it should be made uniform, or turned into a bar graph, or something. More important, this graph emphasizes the results of a single primary study, whereas the article should be emphasizing what reliable reviews say. I suggest removing the graph and basing the coverage of effectiveness on comments like the following instead:

  • "These treatments have been well studied in the short term, with approximately half of children having at least a 50% reduction in seizures after 6 months. Approximately one third will attain >90% reduction in their seizures." Kossoff EH, Rho JM. Ketogenic diets: evidence for short- and long-term efficacy. Neurotherapeutics. 2009 Apr;6(2):406–14. doi:10.1016/j.nurt.2009.01.005. PMID 19332337.

Eubulides (talk) 19:37, 19 October 2009 (UTC)

Graham doesn't like the chart either. I guess it should go. I'm a wee bit concerned you are considering replacing Outcomes paragraphs 2,3,4 with one sentence. I did anticipate that there may be a desire to shorten this section, but not that much. I'll have another read of the paper you cite. I'm a bit concerned about it on two fronts: The abstract doesn't agree with the body: the abstract's "approximately one third will attain >90% reduction in their seizures" becomes "and 24% had a >90% seizure improvement" when you read the body. That's less than a quarter. My second problem is that it doesn't appear to be a meta-analysis (I'm no expert) but a simple arithmetic exercise. Compare this to the meta-analysis performed by Beth Henderson(2006), which admittedly is now out-of-date. I'd appreciate your comments on this. I'll have a go at revising the text here. The important points are (1) the percentages who achieve improvement (>50% reduction), a large improvement (>90% reduction) and who become seizure free; (2) the high drop-out rate; (3) that those who drop out are typically those who failed to see an improvement; (4) these studies are generally for a <12 month period, sometimes just 3. Colin°Talk 20:14, 19 October 2009 (UTC)
  • Update. I've spend some time thinking about the Efficacy section. I'm a strong supporter of just stating facts as facts rather than saying "this study from 2001 found X and another from 2005 found Y". The overall message is summed up by the sentence: "Children with refractory epilepsy are more likely to find the ketogenic diet to be effective than an anticonvulsant drug." which is sourced to a review. Beyond that, however, the details of percentage response and drop out over time are intimately linked to the particular study. There are so many variations in treatment protocol, patient selection and study design and duration. This isn't a pill manufactured on an assembly line. Kossoff's 2009 review on Efficacy discusses the various studies without trying very hard to analyse them together. I think there are three studies that are worth discussing in the article because they have different qualities. The Johns Hopkins study is big and long-term and conducted by the centre that established the protocol used by many other centres since. The Beth Henderson meta-analysis is the biggest such review to date. The Great Ormond Street study is of course notable by being the first and only RCT of any size but was only for three months duration. Colin°Talk 23:16, 27 October 2009 (UTC)
  • Update2. I've moved the review-based efficacy summary to the lead of the Efficacy section. I've also moved the key aspect of this summary to the lead. This leaves the two subsections to discuss trial design and a couple of notable trials. I hope this satisfies the concern that the facts about efficacy should not be sourced to just one study (it is founded on the big meta-analysis). But it also leaves room for the article to discuss trials since most KD review articles discuss the details of notable trials and I've limited this to just the two most significant ones. Colin°Talk 19:58, 31 October 2009 (UTC)

I took another look at the article, and had a go at revising the lead in the hopes of making it easier to follow. One thing that jumps out at me from this go-round (I can't say I've done a real review) is that the boundary between ketogenic diet as a medical treatment, and as a weight-loss strategy, isn't as sharp as the division between this article and Low-carbohydrate diet would imply. Obesity is a medical condition, after all. And even aside from obesity per se, when a person with type 2 diabetes is (perhaps as part of a faddish diet) going ketogenic, there is reasonable speculation that this may help with glycemic control as well.

I suspect it'd be helpful to address this issue more forthrightly, by expanding the Other applications section (currently pretty terse) to talk about medical treatment vs weight loss. While we're on the subject of that section, I had a bit of trouble following the sourcing. Its first 4 sentences have, at the end, citations to Hartman et al. 2007 (PMID 17241206), to Gasior et al. 2006 (PMID 16940764), and to Barañano & Hartman 2008 (PMID 18990309). It's not clear which parts of the text are supported by which of these citations. I was going to fold in a citation to Wylie-Rossett & Davis 2009 (PMID 19793510) on ketogenic diets and diabetes but came up a bit baffled as to how I'd modify the text to do that, given its current sourcing. Eubulides (talk) 06:04, 2 November 2009 (UTC)

I've tidied the source citations a little. Does that help? Another editor added Barañano & Hartman 2008 (PMID 18990309) and I don't currently have access to the full text. There is a little overlap between Hartman 2007 and Gasior 2006, with the latter adding the TBI and stroke. Colin°Talk 20:47, 2 November 2009 (UTC)
I'm not responsible for the Low carbohydrate diet article and don't think it is very good. I agree obesity is a medical condition. Type 2 diabetes mellitus is mentioned already but in no detail. Do you want to have a go at adding something? I see the diabetes sources distinguish between a low carbohydrate diet and a very low carbohydrate ketogenic diet with < 30g carbs (VLCKD). I note that PMID 16018812 considers the VLCKD "an extreme change for most people and therefore we would not recommend it without substantial evidence from clinical trials", though it may be dated. I don't have access to your PMID 19793510 but I have read PMID 18397522, which is an up-to-date article. That article suggests the use of the above-italicised terms and to keep the term ketogenic diet for epilepsy. One distinction is that it is the "very low carbohydrate" aspect that is important. The VLCKD has a hard limit on carbs, but no limit on protein or calories. This is therefore different to the classic KD which is ratio-based, and protein and calorie limited. See also PMID 17684196 and PMID 19099589. I still get the impression that advocates for the LCD or VLCKD have yet to convince the official heart and diabetes bodies that this is a therapy ready for mainstream approval, though they themselves think it "may be one of the most effective dietary treatments for diabetes." Given that much diabetes and obesity research into LCDs spans the spectrum from low to very low carb (and that the ketogenic aspect may be simply a marker for very low carb intake), I think most of the discussion on this topic belongs in Low carbohydrate diet and Diabetic diet. Perhaps soon there will be sufficient material to start very low carbohydrate ketogenic diet. Currently, the only official therapeutic use for a ketogenic diet is for epilepsy. I'm pleased the diabetes researchers feel the need to use a different term to avoid confusion. Colin°Talk 20:47, 2 November 2009 (UTC)
A couple of minor points. Why are the PMIDs in square brackets? Typically they're not, and omitting the brackets will save a bit of screen space. Also, why are the citations to Galen and Hippocrates under Notes rather than References? Eubulides (talk) 06:44, 2 November 2009 (UTC)
See User talk:Fvasconcellos#ketone bodies. Fvasconcellos added the brackets along with moving them after the DOI. I'm tempted to remove the brackets, as our blue text + link icon is sufficient that additional markup [] isn't needed IMO. As for the Galen and Hippocrates texts, please suggest improvements if you have them. I did not refer to the Greek texts so they don't really belong in the References section. So I put those citations in the non-source footnotes and sourced the whole paragraph to Temkin. Colin°Talk 20:47, 2 November 2009 (UTC)

Eubulides, I've partly undone this edit which changed the meaning. Only the classic diet has such strict carb requirements that high-carb foods are "excluded". On the MCT, for example, breakfast could be a Weetabix and MCT emulsion "milk". Plus the change implied the classic diet's 4:1 ratio "provides just enough protein.. and sufficient calories". In fact it is the other way round. The protein and calorie requirements are absolute and children in a fast-growing high-protein-requirements phase may need to use a 3:1 ratio. Lastly, I felt the reverse order made the diet seem "emphasis-based" rather than "technically-based". In other words, driven by some desire to increase or decrease good or bad foods rather than a consequence of achieving a abnormal metabolic state. Perhaps it can still be improved and made easier on the reader. I want to explain these dietary changes are a consequence of trying to achieve ketosis, rather than because some nutritionist thinks we should all eat lots of fat. Colin°Talk 23:00, 2 November 2009 (UTC)

Comments by Maralia (talk · contribs)

I've made some small copyediting tweaks, and left a handful of inline requests for clarification. Other issues:

  1. "The potential use of the diet as a treatment for medical conditions other than epilepsy is, as of 2008, still at the research stage." - this is fuzzy, and the problem originates with the subject 'the potential use'.
  2. "This added weight to conclusions" - 'added weight' is a confusing choice of words in an article about diet.
  3. "In children with refractory epilepsy, the ketogenic diet is more likely to be effective than trying an alternative anticonvulsant drug." - the comparison is diet vs alternative drug; 'trying an' just muddies the waters here. Ditto with the later sentence "Children with refractory epilepsy are more likely to find the ketogenic diet to be effective than trying an alternative anticonvulsant drug."
  4. The Epilepsy section needs work. At present, too much of it reads as a disjointed string of extended definitions, and some of the information appears unnecessary here ("This may affect, for example, the muscles, the senses, the mind, or a combination").
  5. "However, a stubborn 20-30% fail to achieve" - here 'stubborn' unintentionally modifies 'individuals', carried over from the previous sentence.
  6. I find 'carbohydrate' used in many cases where I would say 'carbohydrates'. In the lead, for example, I would not say "if there is very little carbohydrate in the diet", but rather "if there are very few carbohydrates in the diet". Is this an EngVar difference, or something else? In any case, it's comprehendable either way, but I can't tell if it's consistently applied since I don't understand the distinction.
  7. I believe 'fasting' is most commonly understood by the layman to mean complete abstention from food. There are points in the article where it would be helpful to distinguish whether 'fasting' in that instance refers to full or partial fasting.
  8. I find the use of 'fasting' as a transitive verb very strange ("the treatment of seizures by fasting patients was not popular", "began to treat his epilepsy patients by fasting them").
  9. The Trial design section is confusing, beginning with a general statement about efficacy, then darting back and forth between blurbs about the design of early studies and modern ones. Wouldn't the first sentence ("the gauged by...changes in the frequency of seizures") be better placed directly under Efficacy? Is there any way to better organize the contents of Trial design, perhaps by drawing together the two distinct selection bias issues and the modern prospective/intent-to-treat approach?
  10. The sentence "Infants and patients fed via a [[feeding tube|gastrostomy tube]] can also be given a ketogenic diet" makes me wonder: can this not be done via NG tubes as well?
Thanks for an interesting read. It's really quite well done. Maralia (talk) 23:18, 16 October 2009 (UTC)
Thanks for these comments, which are just what I'm looking for, and for taking all that time to carefully read and correct the text. I'll try to work on fixes later. I'll also look into the carbohydrate/carbohydrates issue too (though I see Awadewit has changed many to carbohydrates). Colin°Talk 11:10, 17 October 2009 (UTC)
Re point 7. I'm not sure there are any places where "fasting" does not mean "complete abstention from food". I thought "fasting" was an all-or-nothing activity, with "partial fasting" not really being a fast at all. Of course, water is still consumed. Which parts of the article do you think required clarification. Colin°Talk 22:17, 17 October 2009 (UTC)
Re point 8. I think Awadewit has fixed those. My bad grammar. Colin°Talk 22:17, 17 October 2009 (UTC)
Re point 10. My source says "gastrostomy tube". I'm sure that an NG tube could be used as it certainly is for other enteral feeds. But it is only a short term measure and would be replaced by a gastrostomy tube if tube feeding was required for more than a month or so. An NG tube is not much fun :-( Colin°Talk 22:17, 17 October 2009 (UTC)
Re point 1. I've tried to fix that with this edit. Is it still awkward? Colin°Talk 15:09, 18 October 2009 (UTC)
Re point 2. I've rephrased it with this edit. Colin°Talk 15:22, 18 October 2009 (UTC)
Re points 3 and 5. I've dropped to troublesome words in this edit Colin°Talk 15:25, 18 October 2009 (UTC)
Re point 4. I've had a go at making it less of a lesson here. I'm still struggling with the "This might affect, for example, the muscles, the senses, the mind, or a combination." bit. I'm reluctant to remove all mention of the manifestations of a seizure (most folk probably have never seen one and think only of convulsions) but think that sentence is a bit clumsy so it stands out. It is very hard for me to summarise. Perhaps you guys can help. My source text says "The clinical manifestations of a seizure depend on the specific region and extent of brain involved and may include alteration in motor function, sensation, alertness, perception, autonomic function, or some combination of these." Another source, that lists examples, is I think there is some relevance to mentioning very briefly what a clinical seizure looks like, because we say in the article that the success of the diet is measured by observing and counting seizures, not by some machine monitoring brainwaves. I would like someone reading this article, who perhaps even thought epilepsy was some kind of mental illness, to have some idea of what the diet is treating. Colin°Talk 19:21, 19 October 2009 (UTC)
Re point 9. I've moved the text around as you suggested. Much better. Colin°Talk 19:35, 19 October 2009 (UTC)
Overall, I'm happy with your changes. Regarding fasting, I was thrown by the Hopkins description with 'only fluids' and 'eggnog' (I would call that partial fasting) and by Conklin's long fasts (18–25 days makes me question whether it was full fasting or liquid fasting).
The 'Worldwide variants' section is a good addition but needs a little work. 'Maids' should probably be replaced with the broader and more PC 'domestic workers'. The sentence "Indian parents do not allow their children to fast" is not a great paraphrasing of the original ("a strong cultural aversion exists to fasting children"). Maralia (talk) 05:28, 20 October 2009 (UTC)
Thanks. I've just read the WP article on fasting. I must admit I'd thought the term meant "no food" and that water (or some other zero-nutritional-value drink) would be acceptable. Maybe that's not what most people think of as a "fast". All my sources use the term "fast" without qualification but I'm 100% sure nobody denies these children water. Is there a term for a "no food fast"? The "eggnog" breaks the fast (which is 24hr). I'll review my source here to check the timings and see if the wording could be clearer. Conklin calls his therapy the "water diet" -- see the newspaper insert.
Would "servant" be acceptable rather than "domestic workers", which sounds like something out of a PC job advert than a term anyone would use in conversation. You are right about the "Indian parents" sentence. I found that one hard. When the original text is already so condensed and precise, it is really hard to rephrase. Perhaps Awadewit can help rephrase that piece of information. If not, I'll have another try later. Colin°Talk 12:14, 20 October 2009 (UTC)
Regarding fasting, I think maybe I'm splitting hairs too finely. I don't think anyone would quibble with just saying 'fasting' when describing a water-only fast, as I see now that Conklin meant. Given the restrictions of the diet, though, it is odd that sources describe the Hopkins first-day regimen as allowing 'fluids', unspecified.
For 'maids', I don't like 'servants' any better. I asked a friend who works on books about global issues; she suggested 'hired help'. Maralia (talk) 22:00, 20 October 2009 (UTC)
Comments from Awadewit

I am slowly reading the article! Thanks so much for working on this, Colin!

  • All of the images pass muster for FAC.
  • Reference check:
  • Ref 2 - Why is this a RS?
  1. When the body produces ketone bodies—a state known as ketosis—this has an anticonvulsant effect - This sentence is awkward. Do you mean "When the body produces ketone bodies...seizure rates decline"?
  2. Epileptic seizures occur when cortical neurons fire excessively and/or hypersynchronously, leading to temporary disruption of normal brain function - I've added the word "normal" here, as the brain does not cease to function during a seizure.
  3. Peterman's work, in the 1920s, established the techniques for induction and maintenance of the diet, and documented both positive and negative side effects. - What were the side effects?
  4. In the 1960s, it was discovered that medium-chain triglycerides (MCTs) are much more ketogenic than normal dietary fats (which are mostly long-chain triglycerides) - Can we be more specific than "much more"?
  5. Adults can benefit too, though adherence to the regime becomes more difficult with adolescence. - Why?
  6. The diet is probably less likely to be successful in children with very focal epilepsy (where a single well-defined part of the brain is responsible for the seizures), who are candidates for surgery. - Is the "very" necessary?
  7. Some clinicians consider two dietary variants—the low glycemic index treatment and the modified Atkins diet—to be more appropriate for adolescents and adults. - Why?
  8. In general, I'm curious why the diet doesn't work in adults. Perhaps more could be said about this?
  9. Are there studies on how effective the ketogenic diet is for diseases other than epilepsy?
  10. On admission, only fluids are allowed until dinner, which consists of "eggnog" restricted to one-third of the typical calories for a meal. - Perhaps the eggnog should be described? I was thinking of the Christmas drink with nutmeg and rum.
  11. These spikes are an indication of epileptic activity in the brain, but are below the level that will cause a seizure - Don't these spikes represent seizures in the brain that simply don't manifest physically?

I'll post as I read. Awadewit (talk) 00:12, 17 October 2009 (UTC)

Thanks for the fixes you've already made. I look forward to your comments. I'll try to replace that source. Colin°Talk 11:10, 17 October 2009 (UTC)
Added more comments. Awadewit (talk) 23:37, 18 October 2009 (UTC)
Thanks very much Awadewit, for the comments and copy edits. I'll look at them later. The adult thing is a difficult one but it should be covered better in the article. I think there are physical reasons wrt producing/using ketone bodies, there are practical reasons (young children don't choose their own meals), social issues (peer pressure on adolescence and eating out as an adult), and a fair bit of tradition. I also reckon that most ketogenic diet clinics are set up in paediatric neurology departments or in paediatric hospitals so it is harder to transfer this over to adult medicine.
There isn't so much in the literature on this topic as I'd like but I should be able to find enough. See Eubulides comment on the NICE guidelines -- they don't say why they don't recommend the diet for adults. Part of NICE's remit is to prevent the use (and cost to the taxpayer) of unproven therapies, not just ones that have been proven ineffective. Colin°Talk 08:07, 19 October 2009 (UTC)
Would you consider this a reliable-enough source for commentary on the in-adults issue. The author is the same as appears in many of the papers we cite, but the magazine format relaxes the tone (and in some ways that allows for voicing opinions that might otherwise not make it into a stuffy journal). Anyway, you might be interested in the opinions, even if they don't make the grade for the article. Colin°Talk 13:53, 19 October 2009 (UTC)
That looks like it passes WP:RS - the author is an MD and there is an editor for the site (even a topic editor). Awadewit (talk) 21:34, 21 October 2009 (UTC)
You ask if we can be more specific than "much more ketogenic" wrt MCT oil. I've spent some time investigating this and I'm afraid I can't. Nearly all my sources use exactly the same wording here, with the occasional "considerably more ketogenic". It appears the root source for that fact is:
  • Schön von H, Lippach I, Gelpke W. (1959) Stoffwechselunter suchungen mit einem Mischglycerio der Fettsäuren Mittlerer Kettenlänge. Gastroenterologia 91:199–213.
which I'm not able to read and I suspect few others have either. It is possible that PMID 934725 might offer some detail, but I don't have a copy of that paper.
Ah well. 21:34, 21 October 2009 (UTC)
I've replaced an inferior source on the MCT diet with some better ones in this edit. That MatthewsFriends article contains information about the MCT I've not read anywhere else but it doesn't really make the FA grade. Most papers on the KD are from the US, where the MCT isn't popular. Despite some recent papers on MCT from the London team, and the consensus review giving it equal placing, none of my academic papers describe how an MCT diet is prepared or go into detail of the benefits of "more carbs and protein".
Btw, I'm not saying MatthewsFriends isn't reliable - I just wasn't sure. If you can show it is edited in some fashion or contains the contributions of experts, perhaps it would be fine to use. Awadewit (talk) 21:34, 21 October 2009 (UTC)
You ask some good "why" questions but they are hard to answer from the sources which can be frustratingly terse and dry. I'll keep digging but that's all for tonight. Colin°Talk 20:53, 20 October 2009 (UTC)
Thanks! Awadewit (talk) 21:34, 21 October 2009 (UTC)

I've investigated and, where possible, fixed each of Awadewit's points so far:

  1. Rephrased to Elevated levels of ketone bodies in the blood, a state known as ketosis, lead to a reduction in the frequency of epileptic seizures..
  2. That's fine. Thanks.
  3. I've added the side effects in parenthesis.
  4. See above.
  5. I've dropped this sentence, which was actually contradicted by the cited source. Earlier material says this but current work doesn't and in fact there are several papers boasting success with adolescents, though mainly with those already highly motivated. However, see PMID 18675556 for a contrast where the KD is rather unsuccessful in adults. That recent primary research paper isn't currently cited by any of my reviews. The impression I get is that the success you get with adults and teenagers on the KD depends a lot on the patient, their carers and the ability of the medical staff to motivate them and help create interesting diet choices. If one enjoys food, the restrictions of the KD might seriously impact on quality-of-life and may not be worth it if the gains are minor.
  6. I've reworded this completely as it was wrong: Children with a focal lesion (a single point of brain abnormality causing the epilepsy) who would make suitable candidates for surgery are more likely to achieve good results with surgery than with the ketogenic diet. Should there be a comma in there somewhere? Seems rather a long sentence.
  7. I've added the note that these are "less restrictive". They won't tolerate the restrictions. Although the same researchers will also tell you they can! Hmm. One paper hints that experience has taught them to lower the KD ratio for older patients who find the higher ratio hard to tolerate -- and the MAD/LGIT varients are low (1:1) ratio ketogenic diets. Two things I'd like to know: why don't they consider the MAD for children (perhaps they need the restrictions of the KD and are more sensitive to the additional carbs) and why don't they study the KD for adults who are tube fed (for whom tolerance is not an issue). One paper comments that the KD has potential long-term side effects (stunted growth, cholesterol, bone fractures) that might caution against long-term use in adults -- but then admits that there's no evidence the two other variants are any better here. There's just not enough study data on these two variants to confidently say much.
  8. Above. I've added some adult details to the history.
  9. For diseases other than epilepsy, there is very little one can say. None of the "studies" so far have been of the kind that can establish "effectiveness". Just case reports in a few patients or a pilot study here and there. To say "the KD is effective for" for anything other than epilepsy would be wrong at this stage.
  10. I've added a footnote about the eggnog (which was described later on in the article but now moved to the footnote). My sources put it in "" so I have also. It isn't really eggnog.
  11. The official definition of a seizure is "An epileptic seizure is a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain." So I think it needs to have an observable effect to be a "seizure" (though I have seen the term "subclinical seizure" used). The term "epileptiform activity" seems to be used for activity seen on EEG but not necessarily manifest in the patient. But, yes, I guess they are effectively little "seizures".

Colin°Talk 16:44, 25 October 2009 (UTC)

Comment on Carbohydrate vs Carbohydrates

(moved from above: issue raised by Maralia and some edits made by Awadewit) I'm no grammar expert. Is this a countable/uncountable thing? For example, we don't say "My wife takes more sugars in her coffee than me", or "Add 50g of vegetable oils to the mixture" but would say "I have more marbles than you" and we would say "fructose and glucose are sugars". Same for carbohydrate/fat/protein? Colin°Talk 11:10, 17 October 2009 (UTC)

I think it should be plural. We don't say starch is carbohydrate, we say it is a carbohydrate. These are a class of molecules and there are many different varieties, only a few of which are edible (to humans). Cellulose is probably the most abundant carbohydrate on earth, but we can't eat it. We mainly eat starch and more simpler sugars (plural) such as glucose, fructose, sucrose etc., all of which are carbohydrates. Similarly, "fats" is OK too I think. Graham. Graham Colm Talk 15:05, 17 October 2009 (UTC)

I'm trying to understand this as Awadewit didn't change all the "carbohydrate" to "carbohydrates". A "whole-word" search of my PDF sources produces 766 "carbohydrate" and 232 "carbohydrates". I don't understand when to use one and not the other. We say "high fat diet" or "low carbohydrate diet", not a "high fats diet" or "low carbohydrates diet". Two examples of changes that I don't understand (hence, would probably make the same mistake again):

  1. My source said "in a normal subject by starvation, or a diet containing too low a proportion of carbohydrate and too high a proportion of fat." (American writer in 1921). I wrote "in otherwise healthy people when they were starved or if they consumed a diet that was too low in carbohydrate and too high in fat." This was changed to say "was too low in carbohydrates and too high in fats."
  2. This edit changed two sentences. In the first, "forcing the body to burn fat rather than carbohydrate" was changed to "forcing the body to burn fats rather than carbohydrates". That change seems fine to me. The second, "if there is very little carbohydrate in the diet" was changed to "if there are very few carbohydrates in the diet". This is more problematic to me. Swap "carbohydrate" for "fat" in that example if that helps you see the problem I have. I thought "few" was a counting comparison but "carbohydrates" is plural not because there's more than one molecule but because it refers to more than one kind of carbohydrate (like starch, sucrose and glucose).

What about this example (not in the article): "The low GI diet restricts the dieter to fewer carbohydrates than a normal diet yet it does not contain less carbohydrate than a normal diet, this is because high GI carbohydrates are eliminated." Is that sentence grammatically poor? How would you improve on it?

I'm not trying to be awkward here. Just trying to understand the rules so I can get it right myself. Thanks. Colin°Talk 21:51, 17 October 2009 (UTC)

  • I haven't finished copyediting the article, btw. The source sounds grammatically wrong to me (or the grammar has changed). It would seem to me in the source that carbohydrates should be plural because it is a proportion of a group of different carbohydrates. "Very little" becomes "very few" because carbohydrates can be counted. The last example should read: "fewer carbohydrates" rather than "less" - "fewer" is used for countable things. I hope that helps. Awadewit (talk) 22:56, 18 October 2009 (UTC)
  • I think either is acceptable in some situations, and that where the fact of more than one type of molecule is not at issue (sugar in my coffee), the singular is fine; and when that fact is at issue, the plural must be used. Tony (talk) 02:29, 19 October 2009 (UTC)
Head still hurting but thanks both of you for trying to explain. Perhaps it will sink in later. My GI example was me trying to show that the word can be countable in one sense (Carbohydrates is countable and plural wrt different types of carbohydrates) not not countable in another sense (the amount one eats -- we never say "eat fewer fats", we say "eat less fat"). The source example Awadewit points out as grammatically wrong can be changed to plural without any ambiguity because the "low"/"high" attributes refer to "a proportion of" rather than to the words "carbohydrate" or "fat". Perhaps that's the way to use the plural form people prefer but avoiding any ambiguity that we're counting types of carbohydrates. Colin°Talk 07:52, 19 October 2009 (UTC)
"Fewer" and "less" mean two different things: "fewer" means a smaller number of different carbohydrates, "less" means a smaller total quantity. For example, 20 grams of sucrose, 20 grams of glucose and 20 grams of fructose are less carbohydrates than 40 grams of sucrose and 40 grams of glucose, but the latter are fewer carbohydrates. Obviously the "less" meaning, rather than the "fewer" meaning, is intended here. ___A. di M. 16:08, 19 October 2009 (UTC)
I don't like the sound of "less carbohydrates" for the same reason I wouldn't like "less fats" or "less proteins". But "less carbohydrate, less fat and less protein" all work for me. In your example, I would write "... add up to less carbohydrate than ..." Colin°Talk 17:59, 19 October 2009 (UTC)
OK, having spoken to a dietician, I have changed my mind. "Less plus singular" is probably the better construction. We say (well at least in the UK) that a person can eat too much fat, too much protein and too much carbohydrate, but not "too many fats" etc. Tony was right I think. It is all about context, if the article was about carbohydrates, as a class of organic molecules, then the plural might be preferred—but it is not. No one says "fewer proteins" do they? But we say "Jack Spratt could eat no fat". Graham. Graham Colm Talk 18:56, 19 October 2009 (UTC)
Sorry for all of the confusion! Awadewit (talk) 21:34, 21 October 2009 (UTC)
Comments from GrahamColm (talk · contribs)

Colin, I have made a few edits to the article in the hope of making some points clearer. If I have failed feel free to revert them. As would be expected from you, the sources have been well used. I think the "carbohydrate-carbohydrates" issue is not a big deal—especially since no one has mentioned "protein-proteins"! The article is engaging and well-written but I have a problem interpreting this graph:

Ketogenic diet over 12 months.png

I don't think it is very helpful and I can't really understand it. The colour scheme does not help—they are too similar. Is this just me, or are others struggling with this? When it reaches FAC, alt text will be needed for this, and I would have a problem writing it. Graham Colm Talk 12:51, 18 October 2009 (UTC)

Thanks very much Graham, for all the time you've spent on this (and earlier). Your changes were good. I need to look into the addition of "This is because" wrt disorder of fatty acid oxidation, because porphyria isn't such a disorder. You were right the sentence needed a little work. I'll think about the chart. Perhaps it isn't useful and that study is probably superseded by Neal's 2008 RCT. Colin°Talk 15:30, 18 October 2009 (UTC)
Sorry, I didn't notice that the alt text was already there :( Graham. Graham Colm Talk 16:24, 18 October 2009 (UTC)
Comments from Fvasconcellos (talk · contribs)
  • I've replaced the ketone body structures as requested. Hopefully these are consistent now; they'd be clearer if enlarged, but that will push the gallery length into the next section. I'll also add alt text for these later.
  • The citation style is not completely consistent at present. Would you like citations to be in proper Uniform Requirements style, or do you prefer {{cite journal}} now that Diberri's tool is back online?

More later. Fvasconcellos (t·c) 19:53, 18 October 2009 (UTC)

Thanks, Fvasconcellos. Darn, I thought my citations were fairly well formatted. I'd appreciate if you could fix any issues. If you convert them all to cite journal, I may have to kill you. I stopped using those templates when I couldn't achieve any consistency and when the template-editors seemed determined to invent their own house style but couldn't agree on one The recent citebot problems and issues with bloating hidden meta-data just confirm to me that was a good move. Colin°Talk 20:02, 18 October 2009 (UTC)
Will do. How's this for the images? Fvasconcellos (t·c) 20:22, 18 October 2009 (UTC)
Looks good. Colin°Talk 20:28, 18 October 2009 (UTC)
Excellent. The citations are fairly well formatted, by the way :) Uniform Requirements style follows the NLM's wonderfully exhaustive and esoteric Citing Medicine guidelines, available here. The only real issues are quotation marks/italics where they shouldn't be. Fvasconcellos (t·c) 20:29, 18 October 2009 (UTC)

Fvasconcellos, what do you think about adding the LGIT to the pie chart image? Do you find the chart useful? Unfortunately, the LGIT doesn't have a hard set of figures. My source says "Protein contributes 20–30% of calories, while fat contributes 60%".

Colin°Talk 20:40, 18 October 2009 (UTC)

I do find the chart useful. Is the contribution of carbohydrates to the LGIT 10–20%, then? Fvasconcellos (t·c) 15:33, 21 October 2009 (UTC)
That's the obvious assumption, but the text doesn't say. Colin°Talk 17:04, 21 October 2009 (UTC)
Comments by Mmagdalene722 (talk · contribs)
  • I've made a couple of very minor punctuation and word-order changes to improve flow.
  • This may be fairly minor, but shouldn't the "Epilepsy" section have a link at the top that says, "Main article: Epilepsy"? A lot of the larger articles I've seen do that, but they're mostly for countries...I'm pretty new to Wikipedia, but I thought I'd throw that out there.
  • Is "paediatrician" the correct spelling?
    • I'm not sure about the "Main article: Epilepsy". I think it tends to be used if this was a main topic (e.g. Neurology) and Epilepsy was a sub topic. As for "paediatrician", the article is written in British English so it is fine. See WP:ENGVAR. Wikipedia articles should be consistent but not always in US English. Thanks for helping! Colin°Talk 22:38, 2 November 2009 (UTC)
  • Have you thought about grouping "Initiation," "Maintenance," and "Discontinuation" as sub-categories into one category? They seem to go together, though I don't know what their unifying "header" would be. Just a thought.
  • According to WP:LINK, terms only need to be wikilinked once in an article. I noticed that Johns Hopkins is wikilinked in almost every instance. There are a lot of instances where Johns Hopkins is mentioned, but I didn't want to do it to the whole article if you felt like it was necessary to link it that way. Check the guideline, and if you feel like you need to change the article to conform to it, then go ahead.
  • The Johns Hopkins program is the only one I see mentioned in the article. Is theirs considered the "gold standard" for this treatment? If there are others, I think they should be mentioned.
  • Carbohydrate vs. carbohydrates: not really sure what the British standard is. I've always heard it referenced in the plural unless a specific type of carbohydrate was being referenced; I changed a couple of instances, but then I stopped because I wasn't sure. Look through and see what you think. Sorry if I messed anything up.