1/n Warning: long thread approaching (I should have done it as a blog, but that requires too much effort on LSL). Hi @ChrisPalmerMD and @evolutionarypsy. I haven’t forgotten our Twitter convo on the KD in relation to diet and the gut microbiota...
2/n, but I’ve been having too much fun on leave to get to this. Also, I really didn’t even know where to start! The gut microbiota is a complex topic and there’s a lot we do know, but even more than we don’t.
3/n Straight up, to my knowledge there are no credible studies that have directly tested the impact of a #ketogenicdiet (KD) on the #gutmicrobiota (GM) to date.
4/n We were set to do this last year, using our ‘healthier’ KD (poly/mono fats, focus on plants and fish) but COVID got in the way (we hope to start by end of this year).
5/n However, there are many studies and trials now that are concordant in their findings regarding diet, the GM and health that give us insights, and I’ll touch on these in the thread. Firstly, for general readers, it’s probably useful to know what the GM ‘do’.
6/n Firstly, GM digest, absorb, and metabolise macronutrients that can’t be broken down by human enzymes; this is primarily plant fibre (carbs that are indigestible in the small intestines: called microbial accessible carbohydrates by Sonnenburg et al. – MACs from here on in)
7/n Foods high in MACs include wholegrains, legumes, vegetables, fruits, nuts and seeds (for an excellent review of fibre see nature.com/articles/s4157…).
8/n GM extract bioactive compounds and produce metabolites including short chain fatty acids (SCFAs), which have both local and systemic effects. The role of the SCFAs acetate, propionate and butyrate is becoming clearer; butyrate, in particular, seems to be important in health
9/n GM and the SCFAs they produce influence immune function: In fact roughly 70% of your immune system ‘happens’ in your gut. They also influence drug metabolism & may deal with toxic pollutants.
10/n GM also synthesise vitamin K, vitamin B and CLA, and both synthesise and modulate the production of neurotransmitters such as serotonin, dopamine, epinephrine, norepinephrine, acetylcholine, histamine, and gamma-aminobutyric acid (GABA).
11/n Although these are unlikely to cross the blood brain barrier, they influence gut–brain hormonal communications. GM also influence the metabolization of tryptophan to affect levels of serotonin in the brain.
12/n The metabolites produced by the GM’s breakdown of MACs (and polyphenols) do a lot more than this though. SCFAs directly signal to cells through surface G protein–coupled receptors throughout the body and influence gene expression, including through epigenetic mechanisms.
13/n GM also influence glucose and insulin regulation and likely influence body weight. They also influence the stress response system, and numerous other systems in the body (I don't want to be here all night LOL)
14/n How do we know that they influence mental and brain health? Via numerous lines of evidence. E.g., animal studies now to tell us that as well as influencing the early development of the immune system, early life GM appear to influence brain development.
15/n For example, animals born without a microbiota (germ-free mice) have profound differences in their neurotransmitter systems, stress response systems, brain plasticity, blood brain barrier, immune systems, and behaviour.
16/n New evidence from our team members (e.g. sciencedirect.com/science/articl…)
17/n and ncbi.nlm.nih.gov/pmc/articles/P… suggests that this is also true for humans (we are doing more research in this area).
18/n Also, FMT transplants from humans with mental disorders can transmit the behavioural phenotype of depression (e.g. Zheng et al. Molecular Psychiatry (2016)) and to some extent schizophrenia, with the associated biochemical changes (Zheng et al., Sci. Adv. 2019; 5 : eaau8317)
19/n Recent randomised placebo-controlled human trials have shown supplemental probiotics to reduce mental health symptoms, particularly in those with clinical depression (Liu et al. Neurosci Biobehav Rev. 2019)
20/n Finally, there are consistent differences between people with mental disorders versus controls in their microbiome composition (we have a large SLR under review, covering MDD, SZ and BD).
21/n For a more extensive discussion, see our recent review in Mol Psych nature.com/articles/s4138…, or any of the reviews from @jfcryan and his group.
22/n So, finally onto diet: the evidence base is large now and growing all the time. Many studies comparing the GM of people consuming traditional diets e.g. pnas.org/content/107/33… show that western GMs are different and have far fewer of the bacterial species that produce SCFA
23/n Even short-term dietary interventions that increase plant fibre and reduce animal fats and proteins can effect quite profound changes in GM very quickly ncbi.nlm.nih.gov/pmc/articles/P…
24/n and associated markers of cancer risk (e.g. this cool study where they swapped people's diets for two weeks nature.com/articles/ncomm…)
25/n Based on both animal nature.com/articles/natur…
26/n and human studies (linkinghub.elsevier.com/retrieve/pii/S…) we know that western diets (low in dietary MACs and high in animal fats and proteins) induce a loss of diversity & bacteria that are specialised to breakdown dietary MACs that compounds across generations & is associated w Met dis
27/n Conversely, as documented in observational studies (e.g. pubmed.ncbi.nlm.nih.gov/26416813/) Med-style diets are associated with higher levels of beneficial SCFAs, while animal fat/protein- heavy diets = higher concentrations of problematic compounds such as TMAO
28/n Similarly, healthier (diverse) diets are linked to more diverse gut profiles and lower measures of frailty, including inflammation and cognitive decline (e.g. pubmed.ncbi.nlm.nih.gov/22797518/)
29/n The recent very large and comprehensive PREDICT study is a great read. They looked at healthy or unhealthy foods (animal or plant), diet quality (processed versus unprocessed) and dietary patterns nature.com/articles/s4159…
30/n And fasting and same-meal postprandial cardiometabolic blood marker measurements. They found that ‘microbiome signatures grouped both microbiome & dietary components into health-associated & anti-associated clusters, the latter in agreement with dietary quality & diversity’
31/n Findings from new human intervention studies are concordant with these findings: Med-style diet increases beneficial bacteria and metabolites and is associated with improved measures of frailty gut.bmj.com/content/gutjnl…
32/n This is a really nice recent Med-diet intervention with a lot of detail on the GM and relationships to health: gut.bmj.com/content/gutjnl…
33/n As you might expect, reducing intake of MACs results in reductions in butyrate-producing probiotic species (e.g. ncbi.nlm.nih.gov/pmc/articles/P…)
34/n even Gluten-free (fibre-reduced) diets also seem to reduce beneficial bacteria, in parallel with an increase of E. coli and total Enterobacteriaceae, bacteria typically associated with poor health ncbi.nlm.nih.gov/pmc/articles/P…
35/n And high-protein, reduced carbohydrate weight loss diets appear to result in reductions in cancer-protective metabolites and increased concentrations of detrimental metabolites pubmed.ncbi.nlm.nih.gov/21389180/
36/n There are a few nice reviews on the topic of diet and the gut microbiota including one of ours: pubmed.ncbi.nlm.nih.gov/33693453/
37/n However, as in a lot of medical research in general, and nutrition science in particular, we need more studies. A lot of the studies on the GM and mental health come from animals and we definitely need more dietary interventions in humans.
38/n The studies that I’m most interested in and convinced by are the ones coming from the cancer field increasingly linking dietary intakes and diet-related bacteria to cancer-related biomarkers and treatment outcomes; these are often coming from large human cohorts
39/n For example, the bacterium Bilophila and metabolite TMAO are both linked to increased cancer risk as well as CVD; these are consistently higher in diets higher in animal foods, while fibre-promoted species are associated with better outcomes
40/n For example, De Angelis et al. nature.com/articles/s4159… found that omnivore, vegetarian or vegan diets had an impact on the microbial synthesis of proteins (e.g., flagellin & L-asparaginase)
41/n & metabolites (e.g., butyrate and pyridoxal/pyridoxine) linked to oncogenesis and tumour suppression. More fibre = more flagellin, which has anti-tumor and radioprotective activities and has shown potential in combating tumor growth and radiation-associated tissue damage.
42/n Numerous human studies now link positive immunotherapy and allogeneic stem cell transplant responses in people to (fibre-related) varieties of gut bacteria at baseline; particularly GM diversity, which is consistently correlates with diets high in amount/diversity of fibre.
43/n In summary, so far we consistently see that diets that are high in MACs and low in fat or Med-style diets are associated with or enhance attributes of the GM that are understood to be beneficial, while diets that are low in MACs and/or high in animal proteins/fats
44/n are associated with or enhance attributes of GMand associated metabolites that are believed to be detrimental. But, as usual, we need more studies as there are inter-individual variations in response to food and diet is an insanely complex construct (not to mention the MB!)
46/n In relation to the KD and its impact on the GM in particular, there’s only one credible study that I’m aware of. This was done in children with epilepsy and suggested that a KD helped to normalise a previously ‘dysbiotic’ GM profile in children. ncbi.nlm.nih.gov/pmc/articles/P…
47/n However, this study is confounded of course by the medical condition – does the GM normalise from the diet or from the reduction in fits? This is a common problem in the field – delineating cause and effect.
48/n (Similarly – and VERY importantly – lumping ‘Carbs’ together makes as much sense as lumping ‘Food’ together. Benefits may simply be because people have stopped eating donuts!!)
49/n One of the big issues with the KD is that any benefits are difficult to disentangle from the changes that come with weight loss. The KD can result in weight loss in the short term and this may be responsible for clinical improvements observed
50/n In this sense I think that the Oslo study is really important pubmed.ncbi.nlm.nih.gov/30408717/
51/n To address this issue, they recruited 30 young healthy adults in the recommended BMI range; they were allocated to either a LCHF diet similar to the Atkins diet (max 20g carbs/d), or to control group, who ate their regular diet.
52/n After three weeks LDL-C increases varied between 5-107% (obvs reflecting genetic vulnerabilities) with an average increase of around 45%. This is not good. Mostly strikingly however, was the adverse events reported, which occurred only a few days after the intervention began
53/n Two participants in the LCHF group were removed from the study after one was admitted to hospital with chest pains and severely elevated biomarkers of potential heart damage while the other experienced autoimmune thyroiditis.
54/n Negative cardiac effects are also described in at least one study in children. pubmed.ncbi.nlm.nih.gov/10881264/
55/n The problems associated with the diet from the nutritional perspective (deficiencies) are also well described in the epilepsy literature; thus the necessity for close monitoring and guidance from dietitians.
56/n So from pretty much all the available quality evidence we can say that a traditional (animal-focused) KD is the opposite of a ‘gut-friendly’ diet and that it also carries significant CV and other health risks.
57/n Moreover, although there are some distinct conditions for which #keto is well evidenced (e.g. drug resistant epilepsy), there is a lack of appropriate evidence to support it in psychiatry at this stage. ncbi.nlm.nih.gov/pmc/articles/P…
58/n This is why I would not currently recommend it for anyone until we have the very important safety (first) and efficacy data from properly conducted RCTs.
59/n As I noted, we are planning to test the GM impact of a mono/poly/plant focused KD in healthy people who normally consume a Med-style diet. COVID has stopped us so far, but I’m hoping by later in the year we’ll be able to do it.
60/n Importantly, cognisant of the interest/case studies/and mechanistic studies, and the need for good quality evidence, we ARE planning to assess the possible efficacy of a KD in psychiatry. We’re currently running a pilot study of our ‘healthy’ KD in people with SZ
61/n and are currently trying to get funding (the BIG rate limiter) for an RCT in BD. We also intend to work with @ZSarnyai (and I believe you too @ChrisPalmerMD!) on the human study being planned at JCU (where I am an Adjunct Prof).
62/n So watch this space! (and I look forward to working with you!) 😋
63/n and for all those following along at home, apologies. Go back to what you were doing. I'm now going to bed
• • •
Missing some Tweet in this thread? You can try to
force a refresh
