Chronic diseases are rising despite societal preoccupations with dieting and image consciousness(Reference Elizabeth, Machado and Zinöcker1). Low-carbohydrate diet (LCD), energetic restriction over extended periods and fasting practices, for health benefits, have been an emerging research focus in nutritional epidemiology, health intervention and health promotion(Reference Bolla, Caretto and Laurenzi2,Reference Eissenberg3) . There has been an increasing number of systematic reviews (including a diverse range of meta-analyses of randomised control studies) with a key focus on either LCD or fasting(Reference Santos, Esteves and da Costa Pereira4–Reference Park, Seo and Paek8). However, there have not been any review papers investigating fasting in conjunction with LCD to assess the integrated impact.
Fasting has been practiced and studied for religious, socio-economic, experimental and therapeutic purposes for centuries(Reference Johnstone9). Various forms of fasting have recently become popularised, including alternate-day fasting (ADF) (1-d energetic restriction then at liberty energetic intake the next)(Reference Harris, McGarty and Hutchison6,Reference Park, Seo and Paek8) , 24-h fasting several times per week, daily 16-h fasts(Reference Cho, Hong and Kim5,Reference Welton, Minty and O'Driscoll10,Reference Meng, Zhu and Kord-Varkaneh11) , the Michael Mosley 5:2 (2-d energy restriction and 5 d at liberty) and fasting-mimicking diet(Reference Mosley12) and intermittent fasting/time-restricted feeding (abstaining from food for periods of 16–18 h alternating with normal eating for 6–8 h)(Reference Mattson, Moehl and Ghena13). Some of the fasting methods can be more adjustable to enhance adherence (e.g. intermittent fasting)(Reference Freire14). Fasting research has substantially focused on improving cellular health for longevity(Reference Longo and Panda15), lifespan and healthspan(Reference Eissenberg3). Other significant areas of research into fasting include its impact on different types of cancer(Reference Obrist, Michels and Durand16), inflammatory diseases, neurodegenerative diseases(Reference Mattson, Moehl and Ghena13,Reference Vasconcelos, Yshii and Viel17) , weight control(Reference Cho, Hong and Kim5,Reference Harris, McGarty and Hutchison6,Reference Park, Seo and Paek8,Reference Welton, Minty and O'Driscoll10) , reversing type 2 diabetes(Reference Harris, McGarty and Hutchison6,Reference Johnstone9) , decreasing cardio-metabolic disease risk factors(Reference Cho, Hong and Kim5,Reference Park, Seo and Paek8,Reference Welton, Minty and O'Driscoll10) , improving lipid profiles or dyslipidaemia parameters(Reference Park, Seo and Paek8,Reference Welton, Minty and O'Driscoll10,Reference Meng, Zhu and Kord-Varkaneh11) and decreasing fasting plasma glucose and insulin resistance(Reference Cho, Hong and Kim5,Reference Park, Seo and Paek8) .
There is also a long history of how humans eat that involves restricted, selective or LCD. The popular Palaeolithic diet, based on the pre-agricultural period of eating with a focus on higher protein, fat and selective carbohydrate composition, has seen a re-emergence in the past two decades, from its first introduction in the 1970s(Reference Katz and Meller18). The Atkins, Zone, South Beach and ketogenic diets are also common LCD that are gaining popularity(Reference Bolla, Caretto and Laurenzi2), each focusing on different percentages of carbohydrates, proteins and fats. According to the American Diabetes Association, LCD are defined as having below 130 g/d or 26 % of total energy intake from carbohydrates. Studies on LCD classify carbohydrates into moderate (< 45–40 % energy intake), low (< 40–30 % energy intake) and very LCD (< 30–33 % energy intake)(Reference Fechner, Smeets and Schrauwen19). Systematic review papers on LCD report similar findings to review papers on fasting, indicating a decrease in metabolic syndrome – as a group of risk factors (e.g., blood glucose, blood pressure, TAG, insulin levels, weight gain and an improvement in lipid profiles)(Reference Liu, Wu and Xia7,Reference Fechner, Smeets and Schrauwen19) . Decreasing these risk factors improved chronic lifestyle diseases such as cardiovascular risk factors(Reference Santos, Esteves and da Costa Pereira4), type 2 diabetes(Reference Fan, Di and Chen20), obesity(Reference Freire14) and cardiovascular/cardio-metabolic disorders(Reference Gee and Whaley21).
An LCD is typically considered low in digestible carbohydrates, high in fat and moderate in protein(Reference Freire14,Reference Fan, Di and Chen20,Reference Arbour, Stec and Walker22) . This review considers the term ‘low-carb diet’ as low in starch(Reference Katz and Meller18) and includes indigestible or slowly digested carbohydrates(Reference Svihus and Hervik23) or low glycaemic index and includes thresholds of carbohydrate intake below 45 % energy intake. The general term ‘low-carb’ is used in this study, including ketogenic diets as a low-carb diet. Ketogenic diets include a very minimum carbohydrate (< 20 g/d) and exclude foods with a high glycaemic index, such as simple sugars, highly processed foods, fruits and starches(Reference Freire14). Other LCD may have more flexibility than a ketogenic diet regarding the range of carbohydrate percentages and type of carbohydrate inclusion (fruit, for example is not included in a ketogenic diet, but may be included in a more moderate low carb diet) based on an individual’s health status and physical activity levels(Reference Arbour, Stec and Walker22). However, the foundation of all types of LCD is relatively similar – to decrease the number of carbohydrates consumed and use fat as the body’s preferred energy source. This scoping review aims to explore the evidence on the health outcomes of using an LCD in conjunction with fasting due to the scarcity of reviews around this topic.
Materials and methods
Scoping review design
This study is a scoping review using PRISMA-ScR, which is the Preferred Reporting Items for Systematic Review and Meta-analysis Extension for Scoping Reviews checklist(Reference Tricco, Lillie and Zarin24). This review was performed to explore and systematically map the existing literature to identify concepts, theories and types of evidence and ascertain gaps in research around fasting and LCD in combination(Reference Levac, Colquhoun and O’Brien25).
The review has been conducted by applying the steps outlined by Arksey and O’Malley(Reference Arksey and O’Malley26): (a) identifying and articulating the research question, (b) identifying and implementing the search strategy, (c) selecting relevant studies, (d) extracting and charting data and (e) analysing, summarising and reporting on results.
Identifying the research question
Two independent reviewers (NS, MW) searched for and read articles on fasting and LCD patterns. A consensus was reached for the research question: What are the health-related outcomes of LCD in conjunction with fasting?
Search strategy
Two independent reviewers (NS, MW) conducted a literature search across four databases, Medline, PubMed, CINAHL and Scopus, from inception to the end of 2021. These were chosen as they contain an extensive range of articles from the medical and nutrition field. Searching title and abstracts for the following keywords: ‘low carb*’ OR ‘low-carb*’ OR ‘low glycaemic’ OR ‘low glycemic’ OR ‘low GI’ OR keto* OR paleo* OR Atkins OR zone AND fasting OR ‘intermittent energy restriction’ OR ‘5:2 diet*’ OR ‘16:8 diet*’ OR ‘meal frequency’ OR ‘800 calorie diet*’ OR ‘warrior diet*’ OR ‘eat stop eat’ OR ‘spontaneous meal skipping’ OR ‘caloric restriction’. We limited our searching to the English language and human studies only. There were no limitations on the publication date or age of participants. Figure 1 provides a summary of the search, inclusion and exclusion strategies in the PRISMA flow diagram.
PRISMA flow figure.
Study selection
Papers were screened by title and abstract by all reviewers independently (NS, MW). Results that met the following inclusion criteria were included (1) both fasting and LCD intervention, (2) any health outcome (e.g. treatment and/or prevention of physical, psychological and social aspects of health), (3) quantitative, qualitative and mixed methods research and (4) scholarly peer-reviewed sources in English. Further publications were sourced through snowball searching, including Google Scholar, and scanning reference lists of included papers. A total of fifteen papers were included in the scoping review that met all inclusion criteria.
The main reasons for excluding papers after reading the full text are the key focus on either fasting, or LCD intervention, in isolation. In some studies although both fasting and LCD were used, they were not combined. Furthermore, there were hypothetical studies without any specific intervention and data set. Some examples of the papers excluded are provided below:
Study conducted by Aoki, T. (1981), on ‘Metabolic adaptations to starvation, semistarvation, and carbohydrate restriction’, which focussed mainly on fasting intervention(Reference Aoki27),
Study conducted by Kirk, E. (2009), on ‘Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction’, which focussed mainly on LCD intervention(Reference Kirk, Reeds and Finck28),
Study conducted by Ramakrishnan, T. (1985), on ‘Beneficial effects of fasting and low carbohydrate diet in D-lactic acidosis associated with short-bowel syndrome’ and also study conducted by Nuttall, F.Q. (2015) on ‘Comparison of a carbohydrate-free diet v. fasting on plasma glucose, insulin and glucagon in type 2 diabetes’, which both focussed on either fasting or LCD, and not combining the two(Reference Ramakrishnan and Stokes29,Reference Nuttall, Almokayyad and Gannon30) ,
The study conducted by Brown, A.J. (2007) on ‘Low-carb diets, fasting and euphoria: Is there a link between ketosis and γ-hydroxybutyrate (GHB)’, with no specific intervention (a hypothetical work).(Reference Brown31)
Data extraction, charting and analysing the data
Data extraction was conducted by two independent reviewers (NS, MW) using a predefined format(Reference Aromataris, Fernandez and Godfrey32). Data extraction was done based on (1) author, year, country, (2) research aim, (3) research design, (4) types of fasting and LCD, (5) health outcomes (including health promotion and treatment) and (6) quality appraisal (Table 1). Critical Appraisal Skills Programme (CASP - 2018) checklists were used to assess the quality of the papers based on their designs. No scoring system was provided by CASP. These included various designs, such as observational, longitudinal/cohort (both retrospective and prospective), case study, clinical trial and randomised crossover trial.
Quality appraisal table
Notes: Critical Appraisal Skills Programme. (2018). CASP checklists. https://casp-UKnet/casp-tools-checklists/.
Content analysis was used to create codes, followed by categories and themes. Both authors were engaged in the thematic analysis of the findings separately, and the consensus was achieved via multiple meetings. This resulted in categorising the health-related outcomes of LCD and fasting based on population (children v. adults). The two key themes and subsequent categories were health outcomes of fasting in conjunction with LCD in children (Neurological Conditions – Epilepsy) and health outcomes of fasting in conjunction with LCD in adults (Type 2 diabetes, CVD, inflammatory diseases, weight loss/maintenance and sleep quality).
Results
Study characteristics
The fifteen articles were reviewed with various research designs, including four prospective observational studies, five prospective randomised control trials, one retrospective cross-sectional survey, one retrospective clinical trial, two retrospectives observational and two case study reports. The oldest study was published in 1992, and the more recent studies were published between 2013 and 2021. Studies were conducted in the USA (n 11), Canada (n 1), Japan (n 1), Korea (n 1) and Australia (n 1). Five studies(Reference Kim, Kang and Park33–Reference Hartman, Rubenstein and Kossoff37) were conducted among children aged 2·5–14 years, while the remainder of the studies focussed on adults in the age range of 22–76 years. There was a low number of papers found for this scoping review, particularly control trial studies.
Although a quality appraisal is not necessary for a scoping review, we have assessed the quality of papers for a more thorough analysis of the final included papers. The majority of papers were moderate to high quality (n 10) – this included six high-quality and four moderate-quality papers. There were overall four low-quality papers, including two case studies and two retrospective studies.
The results across the fifteen articles were divided into health outcomes of fasting in conjunction with LCD in children and adults, as these two populations can be different in terms of their response to LCD and fasting, due to their metabolic responses. The only health outcome measured in children was related to epilepsy, as a neurological condition. Various health outcomes were assessed in adult populations, including type 2 diabetes, CVD, inflammatory disease, weight loss/maintenance, sleep quality, insomnia and sleep apnoea (Table 2).
Data extraction of the included papers
LCD: low-carb diets; VLCD: very low-carb diets.
Health outcomes of low-carbohydrate diet in conjunction with fasting in children
Neurological conditions – epilepsy
Five studies compared the impacts of a ketogenic diet on reducing epileptic seizures in children with and without an initial fasting period(Reference Kim, Kang and Park33,Reference Kossoff, Laux and Blackford35,Reference Freeman, Vining and Kossoff36,Reference D’Andrea Meira, Romão and Pires do Prado38) . Although the studies were a combination of high-quality and low-quality designs, they had consistent results, confirming the significant benefits of fasting in conjunction with LCD on decreasing epileptic seizures in children. The studies suggested that although (short-term) fasting in conjunction with LCD provides the ideal results in epilepsy treatment, only focusing on the LCD (usually ketogenic) in this specific population can be more feasible and well-tolerated. A high-quality prospective randomised control trial found significant reductions in seizures after a 3-month follow-up in both the fasting (78 %) and non-fasting groups (67 %). They also reported that 21 % of both the non-fasting and fasting groups were seizure-free after 3 months. The non-fasting group lost less weight and experienced less hypoglycaemia, acidosis and dehydration than the fasting group(Reference Bergqvist, Schall and Gallagher34). The second study, a low-quality retrospective clinical trial, also reported significant reductions in seizures in both the fasting (41 %) and non-fasting groups (42 %). They also found that around 22 % of both the fasting and non-fasting groups were seizure-free at the 3-month follow-up. The non-fasting intervention was better tolerated with less dehydration and fewer hospitalisations.(Reference Kim, Kang and Park33). In the more recent studies (2008–2013), similar results were found – in the low-quality retrospective analysis by Hartman et al., only three of the six participants were able to adhere to the IF/ketogenic protocol for 2 months or longer and seizure improvement ranging between 50 and 99 % seizure reduction was observed(Reference Hartman, Rubenstein and Kossoff37). One participant’s seizures only improved on fasting days. The two participants with the best seizure control reported no adverse effects, with other patients reporting hunger and one losing 1 kg during the regime. In the high quality, blinded crossover study by Freeman et al. which attempted to eliminate ketosis in the crossover group, a decrease in seizures was seen in both groups – 65 % of participants had a greater than 50 % decrease in seizures during the study period with 50 % of those still reporting more than 50 % reduction at 6 months follow-up. This protocol was well tolerated, despite two 36-h fasts within 12 d. There were some adverse effects, with six children reporting vomiting, three reporting fatigue and three reports of hypoglycaemia(Reference Freeman, Vining and Kossoff36). In the study by Kossoff et al., 84 % of participants stated seizure reduction, in those who were fasted seizure reduction, the improvement was faster (median 5 d compared with 14 d without fasting initiation). No difference in fasting v. no fasting was seen at the 6-month follow-up. Over the 6 months, the majority either maintained (n 54) or improved (n 21) their seizure control. Twenty-four children saw a worsening in seizures after an initial improvement(Reference Kossoff, Laux and Blackford35).
Health outcomes of low-carbohydrate diet in conjunction with fasting in adults
Type 2 diabetes
Five studies explored the effect of LCD in conjunction with fasting on health outcomes related to type 2 diabetes(Reference Klein and Wolfe39–Reference Blanco, Khatri and Kifayat43). The majority of these studies have moderate- to high-quality designs (except one case study), and all confirmed the significant results in improvement of the risk factors related to type 2 diabetes, using an LCD in conjunction with fasting. One moderate-quality retrospective cross-sectional study from a fasting and LCD social media group included metabolic data pre- and post-fasting/diet change(Reference Jacobi, Rodin and Erdosi40). They found that 23 % of participants reversed their type 2 diabetes and 21 % reversed their pre-diabetes(Reference Jacobi, Rodin and Erdosi40). In 58 % of participants, diabetes medications were reduced. The HbA1c levels decreased in 54 % of participants. They also reported significant reductions in diagnostic parameters of plasma glucose in 40 % of participants and fasting insulin in 14 % of participants(Reference Jacobi, Rodin and Erdosi40). A second study, a high-quality prospective longitudinal study, reported a reduction in fasting insulin by 24 % after 6 months of combining ADF and LCD but found no change in participants fasting glucose, insulin resistance or HbA1c levels(Reference Kalam, Gabel and Cienfuegos41). A moderate-quality prospective randomised cross-over trial reported decreases in plasma glucose and insulin in the fasting-only protocol but not in the fasting protocol that included intravenous lipid injections(Reference Klein and Wolfe39). Finally, the moderate-quality case study on a normal weight 57-year-old woman with diabetes found that combining LCD and intermittent fasting over 14 months decreased her HbA1c levels from 9·3 % to 5·8 %(Reference Lichtash, Fung and Ostoich42). This was supported by findings from a low-quality case study of a 60-year-old man who self-administered a vegetarian ketogenic diet with intermittent fasting whose HbA1c decreased from 11·5 % to 7 %(Reference Blanco, Khatri and Kifayat43)
Cardiovascular diseases
Four studies with moderate- to high-quality designs focussed on the effect of an LCD and fasting on cardiovascular health and CVD(Reference Klein and Wolfe39–Reference Kalam, Gabel and Cienfuegos41,Reference Bowen, Brindal and James-Martin44) . All papers confirmed the benefits of LCD and fasting in improving the risk factors for CVD. The first study, a moderate-quality retrospective cross-sectional study, found that 56 % of participants’ TAG levels improved and HDL-cholesterol increased for 52 % of participants(Reference Jacobi, Rodin and Erdosi40). They also reported that 71 % of participants were able to discontinue their lipid-lowering medications. The second study, a moderate-quality prospective randomised cross-over trial with two protocol arms separated by 3 weeks, reported a reduced carbohydrate intake was fundamental to the ‘metabolic’ response in short-term fasting rather than overall energetic or energy restriction(Reference Klein and Wolfe39). The third study, a high-quality prospective longitudinal study, reported decreased total and LDL-cholesterol by 8 %, but there were not many changes in HDL-cholesterol, TAG or diastolic blood pressure(Reference Kalam, Gabel and Cienfuegos41). The fourth study (Bowen et al 2018) reported a significant decrease in total cholesterol, LDL, HDL, TAG, C-reacitive protein (CRP) and blood pressure at week 16 in comparison with the baseline with no significant differences between fasting and non-fasting protocols(Reference Bowen, Brindal and James-Martin44).
Inflammatory diseases
Two studies, including one high quality and one moderate-quality design, focused on inflammatory conditions and a combination of fasting and LCD(Reference Jacobi, Rodin and Erdosi40,Reference Manabe, Yoshinaga and Ohira45) . The first, a high-quality prospective randomised control trial, looked at the effects of 6-h fasting without an LCD compared with 18-h fasting and an LCD on patients with suspected cardiac involvement sarcoidosis(Reference Manabe, Yoshinaga and Ohira45). According to the outcomes, higher levels of free fatty acids were detected in the longer fasting, LCD group compared with the shorter fasting group. The second study, a moderate-quality retrospective cross-sectional study, reported the benefits of intermittent fasting and an LCD for other inflammatory conditions(Reference Jacobi, Rodin and Erdosi40). They found that 51 % of participants reported joint pain at baseline, which reduced to 7 % after lifestyle change implementation(Reference Jacobi, Rodin and Erdosi40).
Weight loss/maintenance
Obesity and weight gain, particularly around the waist, are considered the key indicators of metabolic disorders(Reference Liu, Wu and Xia7). Nine papers reported the effect of an LCD and fasting on weight loss and maintenance(Reference Klein and Wolfe39–Reference Kalam, Gabel and Cienfuegos41,Reference Blanco, Khatri and Kifayat43,Reference Bowen, Brindal and James-Martin44,Reference O’Driscoll, Minty and Poirier46–Reference Kalam49) . The majority of these studies had moderate- to high-quality designs, and all confirmed the beneficial impact of an LCD and fasting on weight loss and weight maintenance. A mean weight loss of 16 kg over 3 years was reported by Jacobi et al. (Reference Jacobi, Rodin and Erdosi40) in their moderate-quality retrospective cross-sectional study. Kalam et al. (Reference Kalam, Gabel and Cienfuegos41) and the secondary analyses by Kalam et al. (Reference Kalam47,Reference Kalam, Gabel and Cienfuegos48) reported a net weight loss of 5·5 % of body mass at 3 months and 6·3 % after 6 months in their high-quality prospective longitudinal study. Klein and Wolfe(Reference Klein and Wolfe39) reported weight loss to be greater in the fasting group v. the lipid intervention group in their moderate-quality prospective randomised cross-over trial, although this was attributed to fluid loss. Lichtash et al. (Reference Lichtash, Fung and Ostoich42) reported a weight loss of 4·3 kg over 3 years in their case study of a healthy weight woman with type 2 diabetes. Blanco et al. (Reference Blanco, Khatri and Kifayat43) reported a 9-kg weight loss over 1 year in their low-quality case study. O’Driscoll et al. reported a 9 % weight loss and 8·6 % decrease in BMI and waist circumference(Reference O’Driscoll, Minty and Poirier46). Bowen et al. reported a 10·7 kg weight loss for the fasting (ADF) group and 11·2 kg for the non-fasting group. This result was unexpected, as with energy calculations the fasting group was expected to ingest less energy, leading to a greater weight loss. This result could have been confounded by the inclusion of an eating ad libitum day once per week(Reference Bowen, Brindal and James-Martin44).
Sleep quality
Two studies reported that combining the two dietary patterns did not adversely affect sleep factors(Reference Kalam47). A moderate cross-sectional study by Jacobi et al. (Reference Jacobi, Rodin and Erdosi40) found a 24 % reduction in participants reporting insomnia at the end of the intervention. A moderate-quality longitudinal study by Kalam et al. (Reference Kalam47) found no difference from baseline, 3 months and 6 months in sleep quality, duration, insomnia or the risk of apnoea. That is, sleep quality was not worsened, and sleep duration was not shortened.
Discussion
To our knowledge, this scoping review is the only one to date that has examined the potential health outcomes of LCD in conjunction with fasting. The reason for conducting a scoping review was the paucity of literature and to allow the inclusion of all possible papers on the topic. Overall, the quantity of final included papers is not an issue as far as the review follows a systematic approach(Reference Higgins and Green50,Reference Lang, Edwards and Fleiszer51) . The limited number of papers (particularly control trial studies) indicates the importance of further studies on the topic.
We have categorised our argument on the unique position of fasting in conjunction with LCD into six key sections: (a) the possible advantage of shifting the body fuel from glucose to fat as a more sustainable fuel; (b) improving metabolic syndrome (as the root cause of the most of chronic illnesses); (c) the necessity of individualised approach to the incorporation of fasting in conjunction with LCD; (d) interpretation of blood test markers, particularly lipid package for individuals with LCD; (e) healthy v. unhealthy LCD (well-planned LCD can meet all the Nutrient Reference Values) and finally (f) recommendations for further studies.
LCD in conjunction with fasting (Fig. 2) changes body fuel from glucose to fat, as a more sustainable fuel due to the creation of ketone bodies. This may contribute to glycaemic control (less glucose fluctuation) and hence reduce hunger, which ultimately results in a more viable weight loss and maintenance(Reference Jacobi, Rodin and Erdosi40,Reference Blanco, Khatri and Kifayat43) . According to ‘The carbohydrate-insulin Model of obesity’ (CIM), diets high in carbohydrates (particularly ultra-processed foods) can cause postprandial hyperinsulinaemia, leading to the deposition of energies in fat cells rather than being used as energy in muscle cells. This ultimately results in weight gain due to frequency of eating, reducing metabolic rate or both(Reference Ludwig and Ebbeling52). Kalam et al. (Reference Kalam, Gabel and Cienfuegos41) discussed the benefits of alternative day fasting and LCD as the most clinically significant results to date with a weight loss of 5·5 % of body weight after 3 months(Reference Trepanowski, Kroeger and Barnosky53,Reference Bhutani, Klempel and Kroeger54) , specifically due to carbohydrate restriction to 20–35 %. In addition, combining the two patterns does not impact sleep quality, duration, insomnia or risk factors for obstructive sleep apnoea(Reference Kim, Kang and Park33).
Fasting and Low-Carbohydrate Diet Synergy Flow Chart.
LCD and fasting together provide synergy in reversing metabolic syndrome, as the root cause of chronic diseases (e.g., CVD, type 2 diabetes and inflammation). It will decrease fasting glucose, insulin level and blood pressure, as well as regulate cholesterol by decreasing small dense LDL, decreasing TAG and increasing HDL(Reference Jacobi, Rodin and Erdosi40,Reference Blanco, Khatri and Kifayat43,Reference Manabe, Yoshinaga and Ohira45) . However, it is unclear if improving metabolic syndrome is due to the LCD, fasting or both. One argument is that it is likely that carbohydrate restriction, not the absence of energy, is significant in health outcomes(Reference Klein and Wolfe39). It is argued that the satiating effects of higher concentrations of fat and protein in LCD suppress the appetite, and hence contribute to less fluctuation in glucose and positive health outcomes(Reference Fechner, Smeets and Schrauwen19). It is also argued that although there was a weight reduction, the participants did not report any changes in appetite(Reference Kalam, Gabel and Cienfuegos48).
LCD in conjunction with fasting may not be applicable for some specific populations or demographics (e.g. children, pregnant women, people with anorexia or those with kidney complications). Kim(Reference Kim, Kang and Park33) concluded that the fasting initiation to the LCD for decreasing seizures in children with epilepsy was not therapeutically superior to the non-fasting initiation and was deemed unnecessary as it resulted in more discomfort and distress. The authors discussed that their results were similar to previous case series studies(Reference Vining, Freeman and Ballaban-Gil55,Reference Nordli, Kuroda and Carroll56) in both groups in terms of achieving an appropriate level of urinary ketosis and time to onset of reduction of seizures/being seizure-free. Overall, the benefit of initial fasting on the ketogenic diet is the possibility of screening for metabolic syndrome(Reference Freeman, Vining and Kossoff36). Kossoff states that although fasting is not necessary as long-term outcomes remain similar, fasting can lead to a more rapid reduction in seizures (an average of 5 d with fasting or 14 d without)(Reference Kossoff, Laux and Blackford35). However, this type of initiation may increase the frequency of hypoglycaemia and gastrointestinal problems(Reference Bergqvist, Schall and Gallagher34). It is emphasised that children with epilepsy better tolerate a low-carb diet initiation v. a fasting initiation to the LCD (in which fewer adverse events were reported, such as hypoglycaemia and acidosis). This can be achieved by gradually and briefly increasing the ratio of fat to carbohydrate Kim(Reference Kim, Kang and Park33).
Overall, as childhood is a time of rapid growth and development, any dietary intervention needs to be approached with caution, and in any intervention, care needs to be taken to ingest sufficient energies and protein for continued growth. Children undertaking LCD for epilepsy treatment should be closely medically monitored, particularly in case of utilising any form of fasting(Reference D’Andrea Meira, Romão and Pires do Prado38). Possible adverse effects associated with LCD (either alone or combined with fasting) in children are hypoglycaemia, lethargy, dehydration, metabolic acidosis, gastrointestinal symptoms and weight loss (the most common side effect). For precautionary reasons, multivitamin and mineral supplements can be used if required(Reference D’Andrea Meira, Romão and Pires do Prado38).
It is important to highlight that adherence to LCD and fasting lifestyle patterns can be decided on a personal level, based on health status and weight loss goals, as it may not be achieved easily for everybody(Reference Jacobi, Rodin and Erdosi40). Overall, one diet does not fit all, and optimal nutrition should emphasise high-nutrient density while managing energy balance. Although these two dietary interventions are gaining in popularity, there may not always be significant additional benefits to be gained from fasting(Reference Bowen, Brindal and James-Martin44) or LCD. For many people improving diets by reducing ultra-processed foods (increasing consumption of high-quality protein and fresh fruits and vegetables) can improve inflammation and pain independent of carbohydrate levels(Reference Elizabeth, Machado and Zinöcker1,Reference Field, Pourkazemi and Rooney57) . For example, a positive correlation was found between the risk of metabolic syndrome and higher consumption of carbohydrates, particularly ultra-processed foods in Asian and less developed nations v. non-Asian and more developed countries(Reference Liu, Wu and Xia7). This may be linked to biological and socio-economic status in the study areas and the health literacy around eating patterns.
Individuals need to receive personalised dietary advice, based on their specific situation, which can provide a more sustainable approach, to adhering to the diet. However, there is an argument that the incorporation of more minimally processed foods needs to be considered as a foundation across different dietary patterns. In addition, individual adjustment to the diet may be different based on the health background and socio-demographic status. For example, someone with type 2 diabetes may require more time to adapt in comparison with a healthy individual, due to differences in their levels of insulin resistance. Those with a specific illness require consultation with their health professionals before commencing lifestyle changes. A narrative interpretation of a case study (included in our review) showed that a man with type 2 diabetes self-administered an LCD and fasting regime without professional monitoring. This ultimately resulted in admission to the emergency department with suspected starvation ketoacidosis due to prolonged fasting (5 d of only water)(Reference Blanco, Khatri and Kifayat43). This study concluded that if the fasting period is too long or the energy intake is too low, there is a possibility of ketoacidosis, especially in patients with comorbid illnesses such as type 1 or type 2 diabetes(Reference Blanco, Khatri and Kifayat43). However, how the diet was administered was not reported in this study, and there were no quality control measures, indicating the lack of credibility and relatability of clinical implications which can result in a high level of bias. To our knowledge, there is only limited human research on the dangers of ketoacidosis from low-carbohydrate, low-energetic dietary patterns to date. Overall, we recommend an informed, personalised approach to the prevention, treatment and recovery of each individual (not one size fits all approach). We also suggest providing training in updating healthcare providers’ nutrition literacy (particularly general physicians) based on the most recent and reliable evidence-based research on emerging dietary patterns, including LCD and fasting. Enhancing the nutrition literacy of healthcare providers will impact the way they guide consumers, and subsequently enhance the health and nutrition literacy of consumers to make an informed decision, based on their specific situation.
The other ongoing debate on LCD is related to the lipid profile. It is suggested to interpret lipid profiles differently in LCD, as it can increase the level of both LDL- and HDL-cholesterol but lower TAG (triacylglycerol or triacylglyceride)(Reference Fechner, Smeets and Schrauwen19,Reference Bueno, de Melo and de Oliveira58) , which is not necessarily a negative point. The preservation of HDL-cholesterol could be described by the decrease in postprandial lipaemia (defined as the rise of TAG-rich lipoproteins following food consumption). In addition, elevated LDL in isolation is not a strong indicator of cardiovascular risk(Reference Alique, Luna and Carracedo59–Reference Austin, King and Vranizan61). Increased LDL-cholesterol can be explained by increased levels of larger sized LDL, that are less atherogenic in comparison to small dense LDL(Reference Bueno, de Melo and de Oliveira58). Overall, LDL can be detrimental when joined with fructose and glucose (from high glycaemic index foods), as well as when it is accompanied by a high level of n-6(Reference Alique, Luna and Carracedo59). This can result in the accumulation of glycation/glycated LDL (small dense LDL), and subsequent oxidation, which may not be recognised by the body to be eliminated, and therefore become a risk factor for CVD(Reference Alique, Luna and Carracedo59).
Overall, it is suggested to interpret cholesterol measurements based on the TAG:HDL ratio ≤ 0·8 (TAG ≤ 0·5 (mmol/l); HDL ≥ 1·5 (mmol/l)). Further testing can be done to clarify cardiovascular risk for patients concerned with cholesterol levels such as a Coronary Artery Calcification score, as well as testing subfractions to check for small dense LDL particles(Reference Norwitz and Loh62). LCD improves the lipoprotein profile by improving postabsorptive and postprandial TAG, HDL and distribution of LDL subfractions(Reference Volek and Feinman63). Furthermore, contrary to earlier short-term reports on adverse impacts of LCD, a 10-year prospective longitudinal study shows no cardiovascular risks(Reference Heussinger, Della Marina and Beyerlein64). Interestingly, a 2020 meta-analysis of LCD showed an improvement in cardiovascular risk factors(Reference Dong, Guo and Zhang65). The hypothesis that elevated TC or LDL-cholesterol is the cause of atherosclerosis and CVD has been disproved by numerous studies. Higher levels of LDL-cholesterol have even been shown to be protective in older populations(Reference Ravnskov, de Lorgeril and Diamond66).
It is also paramount to highlight the health benefits of good fat, as a carrier for the fat-soluble vitamins, such as vitamins A, D, E and K. These vitamin deficiencies can result in adverse effects, such as rickets, blindness and haematological disorders. In addition, low consumption of these vitamins can be associated with calcification of the arteries osteoporosis, CVD, type 2 diabetes, depression and some other neurological disorders(Reference Barendse67). Animal-based protein, in particular, organ meats such as the liver, as well as dairy products are valuable sources of these fat-soluble vitamins. When following a well-planned LCD most of the fat consumed will be from animal sources – such as that found in meat, eggs and dairy as well as some sources of plant-based oils such as extra virgin olive oil and nuts(Reference Barendse67).
It is worth noting that just because a diet is low in carbohydrates does not mean it is a healthy diet, as it may still contain many ultra-processed foods. As LCD have grown in popularity, manufacturers have begun to target this market with a variety of low carb or keto foods, which vary widely in quality. LCD increases satiety and decreases libitum energy intake, by increasing the absorption of nutrient-dense foods and replacing ultra-processed foods with more natural and fresh foods(Reference Noakes and Windt68). In addition, LCD improves weight loss/maintenance by improving metabolic changes through enhancing glycaemic control, TAG, HDL-cholesterol and reducing small dense LDL particles that are considered atherogenic(Reference Noakes and Windt68).
One argument often used against LCD is that they differ greatly from the dietary guidelines of many countries and therefore may not enable followers to meet their recommended daily intake of essential nutrients. However, a recent study by Zinn et al. has effectively shown that a well-planned LCD can meet all the Nutrient Reference Values without the inclusion of unusual ingredients such as offal(Reference Zinn, Rush and Johnson69). As with all diets, some planning to meet the Nutrient Reference Values is required, but as the LCD does not exclude any food groups specifically it means a wide variety of nutritious foods can be consumed while following an LCD.
In the end, we have three key recommendations for future studies. First, further high-quality studies (with longitudinal, randomised control trial designs, including higher sample size and longer follow-up) are required regarding the combination of fasting and LCD to enhance the reliability and generalisability of the information. Some of the limitations of the current studies included a small sample size(Reference Klein and Wolfe39,Reference Blanco, Khatri and Kifayat43) ; were self-reported (rather than focusing on active metabolic parameters)(Reference Jacobi, Rodin and Erdosi40,Reference Kalam, Gabel and Cienfuegos41,Reference Manabe, Yoshinaga and Ohira45) ; and had large attrition rates(Reference Kalam, Gabel and Cienfuegos41,Reference O’Driscoll, Minty and Poirier46) , making them susceptible to social desirability bias and underreporting of food intake. There is a need for a design that enhances adherence and compliance, considering this programme as a lifestyle change rather than an unsustainable/short-term diet. Second, there has been a lack of consistency concerning the definition and criteria provided for LCD (different percentages), making the comparison between studies difficult. Macronutrient composition, particularly carbohydrate limits should be clearly defined. In addition, more regular and longer-term follow-ups are required to maintain compliance with the criteria and achieve a more reliable result. We may need to use diverse ranges of classifications of LCD, based on the literature, as different individuals may respond differently to LCD interventions. We recommend classifications of LCD according to Feinman et al. (2015), comprising very-low-carb diet (< 10 % total energy intake or < 50 g/d); low-carb diet (10–26 % total energy intake or 50–130 g/d), and moderate-low carb diet (< 45 % to > 26 % total energy intake or 130–225g/d). We suggest excluding any diet with more than >45% carbohydrate, irrespective of whether it is given a name or not, in the research on the LCD category(Reference Feinman, Pogozelski and Astrup70). Third, we suggest taking into consideration a personalised approach to fasting in conjunction with LCD. These can include individual demographics such as age, illnesses and metabolic conditions.
Conclusion
Two popular dietary lifestyle movements, LCD and fasting, have emerged as a result of common chronic lifestyle diseases, and self-educating, autonomous populations have naturally, or for health reasons, gravitated towards, either/or both patterns of eating. Although this is an emerging field of research, combining the two shows promise of improving and even reversing chronic diseases, possibly unburdening the public health system and enhancing the health span of a large percentage of the world’s population. These eating methods may provide holistic benefits to health and global health economics and the environmental sustainability of eating less often, eating nutrient dense and focusing on functional food choices(Reference Kalam, Gabel and Cienfuegos41).
Our paper has highlighted the multiple benefits of combining an LCD with various fasting protocols. The benefits are multifaceted at three different levels, including prevention (e.g. disease prevention, maintaining or achieving a healthy weight, enhancing sleep quality); diagnosis (e.g. diagnosis of cardiac-related sarcoidosis) and treatment (e.g., improving and reversing type 2 diabetes and prediabetes, CVD, inflammation-reducing joint pain, epilepsy in children). Overall, the combination of LCD and fasting assists in improving the root causes of many diseases, which can be metabolic syndrome.
Although the combination of LCD and fasting can result in synergy, it is still unclear from our work what is the more potent driver of these benefits – is it the fasting or the LCD, or do they have equal importance in improving the disease? There is plenty of literature supporting the use of either fasting or LCD for the same benefits; however, it is not very clear how the combination of both will enhance the synergy. As the first known review of LCD and fasting, we have highlighted the scarcity/paucity of research in an area that is of growing interest and popularity and indicated an area for potential research opportunities. Interestingly, what has been seen in practice is that the two often occur naturally – as patients often report reductions in appetite when following an LCD and will naturally begin to incorporate elements of fasting into their day – in contrast to the need to eat at regular intervals driven by blood sugar fluctuations. Further randomised control trials to confirm the combined benefit are required.
Acknowledgements
This research received no external funding
Both authors contributed significantly to all aspects of the scoping review. N. S. contributed specifically to initiating the research idea, conceptual framework, writing, data analysis and critical argument of the paper. M. W. contributed to searching, data extraction, writing, editing and critical argument of the paper.
The authors declare no conflict of interest.
