Effectiveness of Reduced Carbohydrate Intake

Introduction/Background

The ketogenic diet proposes a reduction of carbohydrate intake, replaced with high fat. Studies convey that low carbohydrate diets promote a higher degree of short term weight loss than conventional low fat diets (Manninen, 2004). Moreover, reductions in fasting blood lipids and insulin concentrations are greater in low carbohydrate diets (Manninen, 2004). Fundamentally, the reduction in carbohydrates renders the body in an efficient metabolic state of dietary ketosis whereby fat is turned into ketone bodies within the liver and burned for energy to utilize in the extra-hepatic tissues. Thereby, short term restriction results in a significant decrease in fat mass and a related increase in lean body mass as fat stores become a primary source of energy (Manninen, 2004). However, low carbohydrate diets may significantly increase fat and cholesterol volume, correspondent with an increase in low density lipoprotein (LDL) cholesterol (Hu et al., 2012). Furthermore, reduction in an accustomed fibre intake may result in constipation, or fatigue induced by a carbohydrate deficiency from altered hormonal states and electrolyte imbalances (Bilsborough & Crowe, 2003). Additionally, complications relative to kidney function may arise (eg. osteoporosis and kidney stones) and can be linked to the long term restriction of carbohydrates (Bilsborough & Crowe, 2003).

Ketosis is a common metabolic adaptation in low carbohydrate diets. Glycogen stores are utilized to meet energy demands of the body when dietary carbohydrates are limited; the reserves are exhausted within 24 to 48 hours of carbohydrate restriction (Bilsborough & Crowe, 2003). However, glycogen is bound to water in a proportion of 1:3g (Bilsborough & Crowe, 2003). Therefore, the subsequent 1-2kg fat reduction can be attributed to diuresis as opposed to burning adipose body fat stores. Consequently, the decrease is not a true indicator of weight loss as glycogen and water stores will be replenished (ie. rebound water weight will be gained) once the diet is terminated due to an influx of carbohydrates, which retain water in the muscle’s glycogen stores (Bilsborough & Crowe, 2003). As depicted in rodent studies, Caton et al. (2009) discerned that the termination of a low carbohydrate diet resulted in weight regain once the habitual diet resumed. Notwithstanding, the diuretic effect is restrained to the first week of the low carbohydrate diet. Subsequent weight loss is entirely due to the ruminant adaption to energy expenditure and balance. Larosa, Fry, Muesing, & Rosing (1980) observed a 7.7 kg loss in participants on the Atkins’ diet in 8 weeks; 1.8 kg lost per week in the initial two week period, and 0.7 kg per week thereafter. Westman et al. (2002) remarked alike results with a range of 0 to 18.6 kg in body weight reduction over 24 weeks (Figure 1). Once glycogen reserves are exhausted, fat oxidation is increased to satisfy the energy demands unfulfilled by gluconeogenesis and triglyceride breakdown as lean mass is inefficiently broken down to glucose and energy deficits proceed uncovered. A directly proportional increase in muscle fatigue and catabolism (conversion of protein to glucose via gluconeogenesis) transpires when muscle glycogen is depleted, however; an important substrate within ATP production (Bilsborough & Crowe, 2003).

Liberation of fatty acids into the blood are oxidized by the liver for energy expenditure to form acetoacetate and further converted to β-Hydroxybutyric acid (ie. ketone bodies) from acetyl CoA, filtered by the kidneys, inducing an increase in renal loss of sodium and consequent water loss. Furthermore, dehydration is common due to the increased water loss associated with ketotic-induced diuresis, onsetting early fatigue in contracting skeletal muscle (Bilsborough & Crowe, 2003).

The long term restriction of carbohydrates pose an increased risk of cardiovascular disease. LDL cholesterol is an eminent factor in atherogenesis, directly correlated with blood β-Hydroxybutyrate (Johnston et al., 2006). Lin & Borer (2016) denote a 30% decrease within physiological insulin resistance 24 hours after three low carbohydrate meals, which increase cardiovascular disease mortality. Moreover, metabolic costs may be associated with the utilization of fatty acids as intermediates of the citric acid cycle imperative for energy expenditure are depleted. Russell & Taegtmeyer (1991) isolated rodent hearts utilizing acetoacetate as an energy source. The researchers observed a 50% reduction in the contractile ability of the heart within an hour; a contractile failure reversed by pyruvate carboxylation.

Indeed, there are benefits and detriments of the dietary regime. However, evidence from clinical and animals trials to achieve a loss in weight and adaptive metabolic risk factors is preliminary.

Review of papers

Research concerning the effectiveness of reduced carbohydrate intake is limited by small sample sizes and short treatment periods. Westman, Yancy, Edman, Tomlin, & Perkins (2002) investigated the effects of a low carbohydrate dietary regime upon body weight and variable metabolic factors in a 6 month trial. 41 overweight (26-33 kg/m^2) yet otherwise healthy volunteers ages 18-65 were assigned to a low carbohydrate diet of <25 g/d with no restricted caloric intake (Westman et al., 2002). Nutritional supplements were provided during the trial period. No formal exercise regime was implemented, however, aerobic exercise was urged. Group meetings were held to reinforce the principles of the diet, and to collect and review measurements. To discern the adherence to diet and exercise, self-report, food records and urinary ketone samplings were obtained (Westman et al., 2002). Moreover, body weight was measured using standard calipers (skinfold thickness in the anterior and posterior upper arm, abdomen, and subscapular skin) at baseline every return visit. Thereafter, the total skinfold thickness measured was employed to estimate body fat composition (Westman et al., 2002).

A mean decrease in body weight of 9.0 +/- 5.3 kg among 39 volunteers was observed (Figure 1). Weight loss correlated with adherence of the dietary regime and ketonuria (P<0.01). Moreover, body fat composition relative to fat mass calculated from skinfold thickness reduced from 36.9 +/- 6.2 kg to 3.0 +/- 5.7 kg (P<0.001; Figure 1). Mean fat mass decreased 2.9% +/- 3.2%; from 42.3% to 39.4% total body weight (P<0.001) (Westman et al., 2002).

Statistically significant changes were obtained for various metabolic parameters relative to changes in serum levels (Table 1). Beneficial effects upon serum lipid levels are indicated; 29 volunteers experienced a net reduction in LDL cholesterol over 6 months (Table 2). Moreover, 37 volunteers had an increase in HDL cholesterol (Westman et al., 2002).

However, there was no objective measure of physical activity, which is potentially confounding. Moreover, adherence is the largest determinant of a regime’s effectiveness. Although group meetings are an objective measure of behaviour adherence, the conceptualization of dietary adherence is disparate, propagated by psychological and socioeconomic determinants. Nonetheless, multiple indicators of adherence to the assigned dietary regime was employed in an attempt to negate the aforementioned issue. Furthermore, all 41 participants developed ketonuria during the trial, strongly correlated with self-reported adherence to the dietary regime. However, it is indispensable to obtain baseline data of macronutrient intake relative to the regime in question to ensure no dietary deficiencies confound the results. Additionally, past dietary intake is principal to document when controlling for baseline, yet macronutrient intake prior to the trial was not assessed.

Often, blood/plasma β-Hydroxybutyrate levels are the only index of ketosis as exhibited in the study of Westman et al. (2002). However, urinary ketones poorly represent the concentrations of blood/plasma and yield less informative results (Table 2). Acetoacetate and acetone are rarely measured and should be investigated in a low carbohydrate dietary regime; direct manipulation is necessary as the correlational approach cannot provide casual evidence of ketones.

Moreover, dietary regime data can be subject to concerns of memory and recall. Additionally, volunteers who completed the dietary records may be more likely to report adhering to the regime. However, macronutrient data was to be recorded within 24 hours of consumption. Also, the usage of skinfold calipers to estimate fat mass poses another limitation. Clasey et al. (1999) discerned that anthropometric estimation yield large mean differences and appreciable inter-individual variability.

Volunteers were not recruited according to strict inclusive criteria, therefore the group is particularly inhomogeneous. However, as the participants were healthy, extrapolating the results to individuals with metabolic diseases should be with discretion. Moreover, no substantial losses to follow-up were incurred as a completion rate of 80% was noted. However, the disadvantageous effects regarding volunteers who did not cohere to the program cannot be eradicated; structured programs are more effective at weight loss than self help approaches (Heska et al., 2003).

Nevertheless, the findings of Westman et al. (2002) emphasize the imminent need for large scale trials on the compound interplay between low carbohydrate diets and long term aftereffects.

The mechanisms and contributing factors underlying the effectiveness of low carbohydrate high fat diets (LC-HFD) remain uncovered. Caton, Yinglong, Burget, Spangler, Tschöp, & Bidlingmaier (2009) examined the effects of a LC-HFD upon body composition and metabolic parameters (eg. growth hormone, IGF-I) in 48 male Wistar rats over a 32 day period.

Two studies were conducted. Study one constituted the maintenance of standard laboratory chow (CH) or LC-HFD in adolescent or mature rodents for 16 days prior to a switch in dietary regime (Caton et al., 2009). However, only mature rodents were maintained on the diets for 16 days in study two in an attempt to illuminate the culmination of LC-HFD upon fat pad mass. All rats were pair-fed to ensure the observations would be due to the macronutrient composition of the diet. Metabolic assessments (eg. energy expenditure) were made at baseline and 16 days post-exposure to the first and second diet with indirect calorimetry (Caton et al., 2009). ANOVA was performed to assess feeding efficiency and corresponding body weight changes relative to age and diet. Moreover, an alpha value of 0.05 was rendered in t-test analysis to examine the disparity between body weight and fat pad mass, with Bonferroni to discern any significant differences between the groups (Caton et al., 2009).

LC-HFD rodents exhibited a significant reduction in body weight irrespective of age and subsequent diet change (Figure 2). Nonetheless, ingesting CH after initial LC-HFD resulted in weight regain in comparison to CH maintained rodents (Caton et al., 2009). Moreover, mature rats maintained on LC-HFD gained remarkably less body weight than CH (CH 27 +/- 1g; LC-HFD 2 +/- 3g; P < 0.01). However, absolute fat pad mass (g) did not notably differ between the diet groups despite a lesser weight gain in LC-HFD. ANOVA revealed decreased IGF-I and growth hormone in the LC-HFD group (922 +/- 60.9 ng/ml; 19.62 +/- 12.26 ng/ml) (P<0.01; P=0.057). Reductions in insulin and glucose concentrations were exhibited (Table 3). Additionally, energy expenditure (EE) corresponded to dietary manipulation. When normalized to body mass, LC-HFD maintained rodents exhibited a reduction in EE among all groups except for LC-HFD first adolescent rodents (Table 4).

LC-HFD may have implications for the alteration of body composition as hormones (eg. GH, IGF-I) known to increase lean body mass diminished within the study; reflective in decreased muscle mass. Declines in IGF-I, lean body mass, and glycogen availability may contribute to the increased fatigue experienced in ketogenic diets. Subsequently, the weight loss procured is not effortlessly sustainable due to an energy imbalance propagating an enhanced drive to regain lost mass.

However, rodent studies are not entirely translatable to human subjects, and moreover, not appropriately designed. Perigonadal fat pads have a large surface area and are readily accessible. Consequently, they are frequently utilized in research, as exhibited in the study of Caton et al. (2009). However, humans do not harbour a fat depot analogous to the fat pads; and thereby cannot be truly deemed as visceral. Furthermore, Bazzano et al. (2014) measured body weight and a myriad of biomarkers in 148 participants on variable carbohydrate diets over a year. The researchers concluded that a low carbohydrate diet was more effective for weight loss and reduced cardiovascular risks. Whereas Vogt (2014) published a conflicting paper; a low carbohydrate diet in maternal rodents alters offspring metabolism whereby risk for obesity is pronounced. There are indeed neuroanatomical similarities between humans and rodents which coincide with food intake and energy homeostasis. However, the study uncovered that the diet damaged the hypothalamus, pivotal for appetite and energy management. No attempt was made to elucidate the contradiction between the reaction of a rodent versus a human, which is misleading. Additionally, rodent strain can determine the susceptibility to diet-induced metabolic changes. If a more resistant strain is utilized, effects may go unnoticed.

Moreover, trials disregard elements of rodent diets (standard laboratory chow; high carbohydrate low fat diet) that have direct metabolic outcomes, such as soy, which has effects akin to estrogen relative to activity, fat storage, and macronutrient and water retention. In contrast, low carbohydrate high fat diets often have sugar as a constituent – associated with weight gain and insulin resistance, ultimately selecting for fat sensitivity. It is notable that a sufficient amount of protein is required to maintain lean body mass, yet the diet in the study of Caton et al. (2009) constituted of low protein. Dietary control in rodents is possible to a degree unfeasible in humans. Metabolic attributes of the human condition cannot be entirely recapitulated in a single animal model.

Summary opinion/Conclusion

A multitude of clinical trials that concern low carbohydrate diets have small sample sizes and insufficient statistical power to detect the incremental changes that occur in metabolic risk factors (Hu et al., 2012). Such factors are important determinants of cardiovascular morbidity and mortality; thereby, it is ineffective to derive conclusions upon the effects of low carbohydrates upon overall health long term. In contrast to the results inferring an increase in cardiovascular risk, Hu et al. propose low carbohydrate diets as an alternative approach for fat mass reduction without worsening metabolic risk factors. Moreover, Bueno, de Melo, de Oliveria, & da Rocha Ataide (2013) denote a doubled average increase in HDL compared to low fat dieters, conferring cardiovascular benefits with an improved cholesterol profile – comparable to Westman et al. (2002) noting an increase in HDL. Nonetheless, six weeks is a short duration of time, and the research conveys that the dietary regime is slightly advantageous in weight loss for up to six months (Fields, Ruddy, Wallace, Shah, & Millstine, 2016). Potential metabolic consequences can be alleviated with increased water, fibre, and calcium intake.

References

Bazzano, L. A., Hu, T., Reynolds, K., Yao, L., Bunol, C., Liu, Y., … & He, J. (2014). Effects of low -carbohydrate and low-fat diets: A randomized trial. Annals of internal medicine, 161(5), 309-318.

Bilsborough, S. A., & Crowe, T. (2003). Low carbohydrate diets: What are the potential short and long term health implications? Asia Pacific Journal of Clinical Nutrition, 12(4), 397-404.

Bueno, N., de Melo, I., de Oliveira, S., & da Rocha Ataide, T. (2013). Very-low-carbohydrate ketogenic diet v. low-fat diet for long-term weight loss: A meta-analysis of randomised controlled trials.

British Journal Of Nutrition, 110(07), 1178-1187. http://dx.doi.org/10.1017/s0007114513000548

Caton, S. J., Yinglong, B., Burget, L., Spangler, L. J., Tschöp, M. H., & Bidlingmaier, M. (2009). Low carbohydrate high fat diets: Regulation of energy balance and body weight regain in rats. Obesity, 17(2), 283-289.

Clasey, J. L., Kanaley, J. A., Wideman, L., Heymsfield, S. B., Teates, C. D., Gutgesell, M. E., … & Weltman, A. (1999). Validity of methods of body composition assessment in young and older men and women. Journal of Applied Physiology, 86(5), 1728-1738.

Fields, H., Ruddy, B., Wallace, M., Shah, A., & Millstine, D. (2016). Are low-carbohydrate diets safe and effective? The Journal Of The American Osteopathic Association, 116(12), 788. http://dx.doi.org/10.7556/jaoa.2016.154

Heshka, S., Anderson, J., Atkinson, R., Greenway, F., Hill, J., & Phinney, S. et al. (2003). Weight loss with self-help compared with a structured commercial program. JAMA, 289(14), 1792. http://dx.doi.org/10.1001/jama.289.14.1792

Hu, T., Mills, K., Yao, L., Demanelis, K., Eloustaz, M., & Yancy, W. et al. (2012). Effects of low- carbohydrate diets versus low-fat diets on metabolic risk factors: A meta-analysis of randomized controlled clinical trials. American Journal Of Epidemiology, 176(suppl 7), S44- S54. http://dx.doi.org/10.1093/aje/kws264 

Johnston, C. S., Tjonn, S. L., Swan, P. D., White, A., Hutchins, H., & Sears, B. (2006). Ketogenic low- carbohydrate diets have no metabolic advantage over nonketogenic low-carbohydrate diets. The American Journal of Clinical Nutrition, 83(5), 1055-1061. 

Larosa, J. C., Fry, A. G., Muesing, R., & Rosing, D. R. (1980). Effects of high-protein, low- carbohydrate dieting on plasma lipoproteins and body weight. Journal of the American Dietetic Association, 77(3), 264-270. 

Lin, P. J., & Borer, K. T. (2016). Third exposure to a reduced carbohydrate meal lowers evening postprandial insulin and GIP responses and HOMA-IR estimate of insulin resistance. PloS one, 11(10), e0165378. 

Manninen, A. (2004). Metabolic effects of the very-low-carbohydrate diets: Misunderstood “villains” of human metabolism. Journal Of The International Society Of Sports Nutrition, 1(2), 7. http://dx.doi.org/10.1186/1550-2783-1-2-7 

Russell 3rd, R. R., & Taegtmeyer, H. (1991). Pyruvate carboxylation prevents the decline in contractile function of rat hearts oxidizing acetoacetate. American Journal of Physiology-Heart and Circulatory Physiology, 261(6), H1756-H1762. 

Vogt, M. C., Paeger, L., Hess, S., Steculorum, S. M., Awazawa, M., Hampel, B., … & Predel, R. (2014). Neonatal insulin action impairs hypothalamic neurocircuit formation in response to maternal high-fat feeding. Cell, 156(3), 495-509. 

Westman, E. C., Yancy, W. S., Edman, J. S., Tomlin, K. F., & Perkins, C. E. (2002). Effect of 6-month adherence to a very low carbohydrate diet program. The American Journal of Medicine, 113(1), 30-36.

Table 1

Effect of a low carbohydrate dietary regime upon metabolic parameters

Table 2

Effect of a low carbohydrate dietary regime upon serum lipid level and 24-hour urinary excretion

Table 3

End-point hormone, glucose, and albumin analysis (study one; mean +/- s.e.m.) in adolescent and mature rodents maintained on CH or LC-HFD for 16 days

Table 4

Energy expenditure (kcal/24 h) normalized for body mass at baseline; 16 days post-maintenance of CH and LC-HFD

Figure 1. The effect of a low carbohydrate diet with additional nutritional supplementation upon body weight (n=41). Fat mass was estimated from skinfold thickness measurement. Fat mass decreased from 36.9 +/- 6.2 kg to 3.0 +/- 5.7 kg. Fat-free mass = body weight – fat mass. The asterisk indicates P<0.001 relative to the comparable change from 0 to 24 weeks. Error bars constitute the standard errors of the mean.            

Figure 2. The development in body weight (g) of adolescent and mature rodents initially maintained on standard laboratory chow (CH) or low carbohydrate high fat diet (LC-HFD) for 16 days prior to a switch in dietary regime (denoted by an arrow) for another 16 days (means +/- s.e.m.). LC-HFD rodents exhibited a significant reduction in body weight compared to CH irrespective of age and subsequent diet change.