Although ketogenic diets debuted in 1920 as a medical strategy to treat refractory epilepsy, its healing properties had not been scientifically evaluated in other medical areas [21, 22]. Recently, its applicability and efficacy are being tested in the treatment of other diseases such as obesity [23], polycystic ovarian syndrome [23], cancer [24], cardiovascular problems [25] and respiratory problems [26]. Multiple worldwide open clinical trials evaluate the tolerability and efficacy of ketogenic diets in the treatment of the aforementioned pathologies as a new medical treatment (http://ClinicalTrials.gov).
KD, understood in a broad sense, refers to any dietetic approximation able to produce a physiological ketosis, this is, an increase of ketone bodies. Classic ketogenic diets [21], fasting periods [27], time restricted feeding [13], caloric restriction diets [28], or intense physical exercise [29] constitute different strategies to produce ketonemia (increase of main ketones bodies, acetoacetate and β-hydroxybutyrate, in blood). Moreover, supplements that mimic the ketosis state as ketone esters [30] or ketone salts [31] have been developed in an attempt to overcome disadvantages of KD without a modification of the diet.
However, little is known about the underlying mechanisms of action of KD. The most accepted hypotheses point out: metabolic changes, alteration of the signalling pathways, changes in the production of hormones and neurotransmitters, epigenetic modifications [22, 24, 27], and as would be expected, modulation of the microbiota [32].
As we mentioned above, the gut microbiota plays an intermediary role between diet and host physiology. Diet affects composition, diversity and functionality of the gut microbiota and these changes in the gut microbiota are inducible and reproducible [33].
Currently, few data is available about the effects of ketogenic diets on gut microbiota composition [34]. (In this review, we have used the term “keto microbiota” to define a profile of gut microbiota moulded by a keto diet). Most of existing data about the KD impact into microbiota comes from epilepsy studies. Recent studies have reported that gut microbiota changes induced by a KD are required to improve the symptomatology of some diseases such as autism [35], epilepsy [33], or sclerosis [36].
3.1 Ketogenic diet, keto microbiota and epilepsy
The classic ketogenic diet (CKD) is a high-fat, adequate-protein, low-carbohydrate diet [21]. The most common ratio in this diet is 3:1 or 4:1. That is, 80–90% of the energy comes from fat and 10–20% from the combination of carbohydrates and proteins [24]. The term was coined by Wilder in 1921 who found that fasting caused an improvement in their epileptic patients and tried to mimetic the ketosis state provoked by fasting with a very low carb diet [37]. Since then, CKD has been the treatment of choice in epileptic refractory patients [21].
Gut microbiota profile is significantly different between healthy and epileptic individuals. KD treatment is able to reshape gut microbiota in humans and rodents [38, 39]; and this keto microbiota is required to avoid seizures. In fact, mouse models of refractory epilepsy showed that those given antibiotics or reared in a germ-free environment were resistant to seizure protection from KD, while keto microbiota fecal transplant helped mice with seizure control. Therefore, these results support that keto microbiota is necessary to protect against seizures [32, 40].
Interestingly, after a KD intervention, patients were differentiated into responder or non-responder subjects according to their gut microbiota changes, suggesting that the effectiveness of a KD was driven by the gut microbiota [5, 32]. Moreover, responder and non- responder groups differed in gut bacteria profiles at the level of order, family and genus, but also in microbial metabolites production. Such bacterial metabolites could be act by restricting precursors availability to synthetize inhibitory neurotransmitters involved in seizure control [40].
In parallel, Hampton et al. revealed that certain combinations of bacteria are required to improve epileptic symptomatology, for instance: they exposed that “Akkermansia muciniphila and Parabacteroides sp. colonization together but not alone protected against seizures in germ-free mice fed the ketogenic diet” [40]. Taken together, these findings underlie that microbiota is a complex system, where interactions between different species enable generate determined profiles of metabolites responsible to provoke a physiological response in host.
By contrast, in spite of benefits of keto microbiota in a growing number of diseases, Tagliabue reported that prolonging the KD for 3 months could cause dysmicrobism with damage to the gut health [41]. Consequently, they recommended prebiotics or probiotics treatment to re-establish gut microbiota and intestine homeostasis [41]. However, more follow-up studies are required in order to monitor the changes of the microbiota profiles with KD, and this highlights the necessity to monitor side effects and take into account possible dysbiosis.