BCAA and T2D

Evidence from cross-sectional and longitudinal research has shown that elevated plasma BCAA concentrations in humans are strongly associated with an increased risk of IR and T2D^{(Reference Arany and Neinast12)}. One meta-analysis suggests that for every one SD increase in plasma BCAA concentrations, the relative risk of developing T2D is 1·35 (95% CI: 1·25, 1·45)^{(Reference Lotta, Scott and Sharp13)}. The BCAA are leucine, isoleucine and valine, and get their name from their hydrophobic side chains that have branched methyl groups. They are classed as essential amino acids because they cannot be synthesised by humans. They must be obtained from dietary sources and play a critical and varied role in human physiology^{(Reference Bender14)}. BCAA are involved in the promotion of muscle protein synthesis^{(Reference Fuchs, Hermans and Holwerda15)}, down-regulation of muscle protein breakdown^{(Reference Louard, Barrett and Gelfand16)}, regulation of energy metabolism^{(Reference Cummings, Williams and Kasza17)}, cellular signalling^{(Reference He, Gong and Zhang18)}, as well as emerging roles such as immune system support and influencing brain neurotransmitters^{(Reference Monirujjaman and Ferdouse19)}. Fig. 1 offers a schematic representation of the varied known and emerging metabolic and physiological roles of BCAA.

Figure 1. Emerging metabolic and physiological roles of BCAA. Figure based on the review by Monirujjaman and Ferdouse^{(Reference Monirujjaman and Ferdouse19)}. BCAA, branched-chain amino acids; T2D, type 2 diabetes.

Circulating fasting plasma concentrations of BCAA are under the tight homeostatic control of BCAA catabolism enzymes, which respond to variations in nutrition, exercise and hormones. In the metabolically healthy, fasting plasma concentrations of BCAA are around 100 µM for leucine, 200 µM valine and 60 µM isoleucine^{(Reference Neinast, Murashige and Arany20)}. Nutritional intake of foods high in BCAA (see Table 1) typically elevate plasma BCAA concentrations postprandially, which then up-regulate the BCAA catabolic pathway, bringing concentrations back to baseline typically within 3–5 h of eating^{(Reference Hagve, Simbo and Ruebush21)}. Exercise also increases the flux of BCAA oxidation, with concentrations decreasing during exercise as the muscle uses BCAA for fuel, returning to pre-exercise concentrations during the recovery period because of increased muscle protein turnover^{(Reference Gawedzka, Grandys and Duda22)}. Fluctuations in certain hormones also influence BCAA homeostasis, with insulin, glucocorticoids, thyroid and female sex hormones known to affect key BCAA catabolic enzymes that either up-regulate or down-regulate BCAA oxidation to maintain typical concentrations^{(Reference Shimomura, Obayashi and Murakami23)}.

Table 1. Total branched-chain amino acid content of common supplements and foods

g, grams.

* Cooked portion weight

† Portion size from British Nutrition Foundation⁽²⁴⁾

‡ Portion size from British Dietetic Association⁽²⁵⁾

^§ Portion size from Optimum Nutrition^{(Reference De Bandt, Coumoul and Barouki26–Reference de Hart, Mahmassani and Reidy30)}

It has been known since the 1960s^{(Reference Felig, Marliss and Cahill31)} that individuals with obesity and IR or T2D have elevated fasting plasma BCAA concentrations, but there was little research interest in exploring their potential role in the aetiology of IR and T2D until the seminal work of Newgard et al. ^{(Reference Newgard, An and Bain32)} sparked a revitalisation of interest in the topic. Newgard and colleagues identified an elevated BCAA ‘signature’ in obese human adults and then demonstrated in rats that a high fat diet in combination with supplementary BCAA directly contributed to the development of IR associated with obesity. Obesity related increases in human fasting plasma BCAA concentrations have been reported to range from around 10%^{(Reference Connelly, Wolak-Dinsmore and Dullaart33)} up to around 25%^{(Reference Lynch and Adams34)} in comparison with lean individuals.

Elevated fasting plasma BCAA are prognostic of T2D. Wang et al. ^{(Reference Wang, Larson and Vasan35)} investigated 2422 normoglycaemic adults as part of a nested case–control study from the Framingham Offspring cohort. They found highly significant associations between baseline fasting plasma BCAA and future risk of T2D with those in the upper quartile of plasma BCAA concentrations (over 481·3 µM) being at least four times more likely to develop T2D during the 12-year follow up period. A systematic review of twenty-seven cross-sectional and nineteen prospective studies, which included Wang et al. ^{(Reference Wang, Larson and Vasan35)}, found that each SD increase in BCAA concentrations equated to a 36% increased risk of developing T2D in the future^{(Reference Guasch-Ferre, Hruby and Toledo36)}. Lee et al. ^{(Reference Lee, Watkins and Lorenzo37)} examined the data of 685 adults of differing ethnicities from the Insulin Resistance Atherosclerosis Study and found that for each SD increase in plasma fasting BCAA concentrations, there was an associated two-fold increase in the odds ratio of developing T2D at the 5-year follow up in Caucasians and Hispanics but not in African Americans. The authors suggested that one explanation for the difference between ethnicities could be because BCAA concentrations are influenced by genetic factors^{(Reference Shah, Hauser and Bain38)}, but further research was needed. In a cross-sectional cohort of sixty-nine healthy children, McCormack et al. ^{(Reference McCormack, Shaham and McCarthy39)} found that children aged 8–13 years with obesity already had elevated plasma BCAA concentrations and that baseline concentrations were moderately positively correlated (coefficient 0·44) with IR at the 18-month follow-up visit. By contrast, there have been conflicting results from both animal and human intervention studies that have attempted to explore whether the relationship between elevated plasma BCAA concentrations and T2D is causal or if BCAA are simply co-incidental early biomarkers of impaired insulin action. However, there is currently no consensus opinion^{(Reference Ding, Wang and Lu40)}. The main hypotheses include persistent activation of the mammalian target of rapamycin (mTOR) signalling pathway and dysregulated BCAA catabolism and are outlined here.

BCAA stimulate the mTOR signalling pathway, which coordinates regulation of metabolism and cell growth in all human tissues^{(Reference Saxton and Sabatini41)} and regulates messenger RNA (mRNA) translation which leads to an increase in muscle protein synthesis^{(Reference Vary and Lynch42)}. Animal studies suggest that persistent activation of the mTOR pathway by dietary BCAA induces IR by inhibiting the action of cellular insulin receptors^{(Reference Tremblay, Krebs and Dombrowski43,Reference Manning44)} (Fig. 2). However, human studies have not consistently found that diets high in BCAA have this effect. Weickert et al. ^{(Reference Weickert, Roden and Isken45)} reported that a diet with 25–30% of total energy from protein from low fat meat and dairy and a daily 58 g mixed whey and pea protein supplement, had only a transient effect on the expression of S6K1 (a key downstream signalling protein in the mTOR pathway)^{(Reference Um, D’Alessio and Thomas46)} in the adipose tissue of the overweight adult participants. Although IR (measured using the euglycaemic–hyperinsulinaemic clamp procedure) was found to increase at the 6-week time point, after 18 weeks of the high protein diet, the change in insulin sensitivity was no longer significant. The authors speculated this could be due to the observed adaptations to the diet in participants, which was associated with a reduction in their abdominal fat and increase in lean body mass.

Figure 2. Proposed role of BCAA in persistent activation of mTORC1 leading to insulin resistance. Adapted from Chen and Yang^{(Reference Chen and Yang47)} and Lynch and Adams^{(Reference Lynch and Adams34)}. BCAA, branched-chain amino acids; IRS1, insulin receptor substrate 1; mTORC1, mammalian target of rapamycin complex 1; S6K1, S6 kinase beta 1.

Magkos et al. ^{(Reference Magkos, Bradley and Schweitzer48)} examined BCAA, mTOR activity and insulin sensitivity in an obese cohort who lost body weight following gastric bypass or gastric band surgery. They found that plasma BCAA significantly decreased, whilst skeletal muscle insulin sensitivity increased. However, there was no change in the rate of skeletal muscle mTOR phosphorylation, implying that the mTOR pathway was not the mechanism that induces IR. Smith et al. ^{(Reference Smith, Yoshino and Stromsdorfer49)} also demonstrated in obese post-menopausal women that whilst WP ingestion (0·6 g/kg fat free mass, given during a 240-min hyperinsulinaemic–euglycaemic clamp procedure) does play a regulatory role in postprandial glucose homeostasis and IR, this was independent of the mTOR pathway. In support of these findings, Maida et al. ^{(Reference Maida, Chan and Sjøberg50)} demonstrated that whilst manipulation of dietary BCAA did impact mTOR activation, the extent of activation (either reduced or increased) did not impact IR in murine models of obesity and T2D.

An alternative hypothesis is that high concentrations of plasma BCAA are driven by dysfunction in the BCAA catabolism pathway. Unlike other amino acids which are catabolised in the liver, BCAA escape the splanchnic territory and are catabolised mostly in skeletal muscle, but also in adipose and other tissues^{(Reference Bender14)}. The first step in the BCAA catabolism pathway involves a reversible transamination, catalysed by branched-chain amino acid transferase (BCAT) which produces branched-chain keto acids (BCKA). The BCKA are then irreversibly decarboxylated by branched-chain a-ketoacid dehydrogenase (BCKDH) to produce branched-chain acyl-CoA esters. The final step of BCAA catabolism involves three separate multi-stage pathways for each BCAA, leading to the production of either acetoacetate, acetyl-CoA or succinyl-CoA^{(Reference Dimou, Tsimihodimos and Bairaktari51)}. An overview of BCAA catabolism is presented in Fig. 3.

Figure 3. Overview of BCAA catabolism. BCAT and BCKDH are enzymes catalysing the first two steps of BCAA catabolism with the end products entering the Krebs cycle. BCKDH is activated via dephosphorylation by PPM1K and deactivated via phosphorylation by BCKDK. Dashed lines indicate multi-stage catabolic pathways. Adapted from Arany and Neinast^{(Reference Arany and Neinast12)} and Dimou et al. ^{(Reference Dimou, Tsimihodimos and Bairaktari51)}. BCAA, branched-chain amino acids; BCAT, branched-chain amino acid transferase; BCKDH, branched-chain keto-acid dehydrogenase; BCKDK, branched-chain keto-acid dehydrogenase kinase; P, phosphate; PPM1K, mitochondrial protein phosphatase 1K.

She et al. ^{(Reference She, Van Horn and Reid52)} examined changes in both the BCAT and BCKDH enzymes in the tissues of obese rodents, and in tissue from obese humans before and 17 months after bariatric surgery. They found obesity-related tissue-specific changes in BCAT and BCKDH levels, both in the rodent models of obesity compared with the lean controls, and in the human tissue samples collected before and after bariatric surgery. BCAT and BCKDH were down-regulated in adipose tissue of obese rodents, but not in liver or muscle. BCAT and BCKDH were up-regulated in the human adipose tissue following bariatric surgery. The authors speculated that excess adiposity and impaired oxidation of BCAA in adipose tissue may play a previously underestimated but important role in elevating fasting BCAA concentrations. Conversely, later work by Neinast et al. ^{(Reference Neinast, Jang and Hui53)} suggested that even in models of obesity, BCAA oxidation in adipose tissue only accounted for around 5% of whole-body BCAA oxidation, shedding doubt on the role of suppressed BCAA oxidation in adipose tissue as the sole cause of elevated plasma BCAA.

BCKDH is regulated by post-translational modifications, meaning its activity is controlled after the enzyme is made. BCKA dehydrogenase kinase (BCKDK) adds a phosphate group to BCKDH (phosphorylation) which inhibits its function, leading to an increase in BCAA and BCKA concentrations. By contrast, mitochondrial protein phosphatase 1K (PPM1K) removes this phosphate group (dephosphorylation) which activates BCKDH leading to a decrease in BCAA and BCKA concentrations. (Fig. 4). Several studies have shown that BCKDK is overexpressed and PPM1K underexpressed, leading to elevated BCAA and BCKA in animal models of obesity and diabetes^{(Reference Wada, Kobayashi and Kohno11,Reference White, McGarrah and Grimsrud54,Reference Zhou, Shao and Wu55)} . Pharmacological interventions that target BCKDK have also provided valuable evidence by demonstrating that suppression of this enzyme leads to reduced plasma BCAA and BCKA and attenuates IR in obese mice models^{(Reference Zhou, Shao and Wu55)}. Bai et al. ^{(Reference Bai, Long and Song56)} added further evidence for the role of BCKDK/PPM1K dysregulation in elevating BCAA and inducing IR when testing high doses of the drug rosuvastatin in mice. Rosuvastatin is commonly prescribed to lower plasma/serum cholesterol but has been linked to an increased risk of developing T2D in humans. Over a 12-week period, high doses of the drug induced a diabetic state in mice along with a 20% elevation in serum BCAA. There was up-regulation of the mRNA expression of BCKDK, and down-regulation of the expression of PPM1K in white adipose tissue and skeletal muscle, indicating that rosuvastatin affects the expression of these enzymes, which then inhibits the action of BCKDH, directly causing elevated BCAA and BCKA concentrations.

Figure 4. Flow diagram of literature search and selection. BCAA, branched-chain amino acids; T2D, type 2 diabetes; WPC, whey protein concentrate; WPH, whey protein hydrolysate; WPI, whey protein isolate.

Several other mechanisms have been proposed to contribute to elevated fasting plasma BCAA. Gut microbiota in insulin-resistant individuals have been shown to have increased BCAA biosynthesis and gut BCAA transport potential, which are associated with increased plasma BCAA concentrations^{(Reference Pedersen, Gudmundsdottir and Nielsen57)}. Certain metabolites of BCAA catabolism have been implicated in promoting insulin resistance including C3/C5 acylcarnitines^{(Reference Rousseau, Guénard and Garneau58)} and 3-hydroxy-isobutyrate (HIB)^{(Reference Bishop, Machate and Henning59)}. Brown adipose tissue has been shown to be involved in clearing BCAA and contributing to metabolic health^{(Reference Yoneshiro, Wang and Tajima60)}. It is beyond the scope of this review to cover all of these proposed mechanisms in detail. For further discussion of the pathophysiology and potential underlying mechanisms involved, including those mentioned here, there are several recently published comprehensive reviews on the topic^{(Reference De Bandt, Coumoul and Barouki26,Reference Ding, Wang and Lu40,Reference Vanweert, Schrauwen and Phielix61–Reference Abdualkader, Karwi and Lopaschuk63)} .

Impact of dietary intake of BCAA

Dietary intake of BCAA varies significantly depending on the type of food consumed, as different foods have variable amounts of BCAA (Table 1 provides an overview of the BCAA content in a range of common foods and supplements). However, the impact of dietary BCAA intake on elevated plasma BCAA remains controversial. It is not clear whether specific foods high in BCAA, an overall diet high in BCAA from different sources, or the metabolic state of an individual plays the greatest role in determining fasting plasma concentrations^{(Reference Merz, Frommherz and Rist64)}. In animal studies, She et al. ^{(Reference She, Olson and Kadota65)} reported that plasma BCAA and BCKA, measured using metabolomics, were elevated by 45–69% in obese Zucker rats fed a standard chow diet ad libitum compared with the lean control rats. By contrast, Roquetto et al. ^{(Reference Roquetto, Moura and de Almeida Santos-Junior66)} fed mice either a high-fat diet or low-fat diet with 13% of total energy intake from protein provided by WP or WPH for 16 weeks but found no elevations in plasma BCAA when compared with the casein control. In a meta-analysis of rodent studies by Solon-Biet et al.^{(Reference Solon-Biet, Griffiths and Fosh67)}, there was a positive association between dietary BCAA intake and plasma BCAA concentration. However, the animal’s existing intake of BCAA was important as to whether a dietary increase in BCAA could influence plasma BCAA. Animals on previously low BCAA diets showed the biggest increases in plasma BCAA, whereas there was little change in animals already on high BCAA diets. Interestingly, in both this meta-analysis^{(Reference Solon-Biet, Griffiths and Fosh67)} and the author’s previous work^{(Reference Solon-Biet, Cogger and Pulpitel68)} there was a plateau effect of dietary BCAA on circulating plasma levels in mice with a maximum plasma level of 40 µg/ml BCAA observed no matter how much further dietary intakes were increased. Whether there is a similar plateau effect in humans remains to be seen, especially since BCAA are metabolised very differently in humans compared with rodents and even other primates^{(Reference Suryawan, Hawes and Harris69)}.

In humans, Woo et al. ^{(Reference Woo, Yang and Hsu70)} found supplementing twelve obese prediabetic participants with 20 g of BCAA (10 g leucine, 5 g isoleucine, 5 g valine) a day for 4 weeks had no effect on their fasting BCAA concentrations, but did result in lower plasma glucose during an oral glucose tolerance test at the end of the study compared with baseline. However, in a 6-month intervention in elderly males with T2D, supplementation with 2·5 g leucine three times a day (7·5 g in total), increased fasting plasma leucine concentrations by 13% whilst simultaneously decreasing concentrations of isoleucine and valine by 16% and 23%, respectively, compared with the controls. However, the leucine supplementation did not change any measures of glycaemic control, which included basal fasting insulin and fasting blood glucose concentrations^{(Reference Leenders, Verdijk and van der Hoeven71)}. This study offers good evidence that dietary intake influences fasting BCAA concentrations but also suggests there may be a distinct metabolic response to supplementation with leucine alone compared with mixtures of all three BCAA.

Elshorbagy et al. ^{(Reference Elshorbagy, Jerneren and Basta72)} showed that thirty-six overweight participants who removed meat, poultry, eggs and dairy products (foods high in BCAA; Table 1) from their diets for 6 weeks as part of a religious fast, had a 15% decrease in fasting plasma BCAA concentrations (leucine −14%, isoleucine −12% and valine −20%) within 1 week of commencing the intervention, which was maintained until the end of the 6-week intervention. There were no changes in body weight and no changes in fasting plasma insulin or glucose concentrations. In a pilot study in six healthy participants placed on a diet where BCAA were restricted by 75% compared with the control diet, circulating plasma BCAA concentrations dropped by 50% (from 437 ± 60 to 217 ± 40 µmol/L) in just 7 d, compared with the control participants on iso-nitrogenous, iso-energetic diets. There was a marginal improvement in insulin sensitivity as measured by the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) in the BCAA-restricted group^{(Reference Ramzan, Taylor and Phillips73)}. Using data from the Karlsruhe Metabolomics and Nutrition study, Merz et al. ^{(Reference Merz, Frommherz and Rist64)} found that an overall ‘unhealthy’ dietary pattern, high in saturated fat that included meat, sausages, sauces, eggs and ice cream, and low in fibre, could explain 32% of the variation in plasma BCAA of participants. The authors speculated that the high saturated fat and low fibre intakes could be responsible for the increased risk of T2D rather than the elevated BCAA. However, as a cross-sectional study they were not able to investigate any causal relationship between diet and BCAA levels. Understanding of the extent to which dietary patterns, individual foods or individual nutrients play a role in elevated BCAA and the subsequent aetiology of IR and T2D is still evolving.

Whey proteins

WP are constituents of dairy milk, comprising approximately 20% of the total protein content. WP contain all twenty amino acids^{(Reference Almeida, Alvares and Costa9)} and are the most concentrated source of dietary BCAA available (Table 1). They comprise an array of protein sub-fractions and peptides with a remarkable range of beneficial properties, including improving markers of glycaemic control^{(Reference Chiang, Liu and Loh74)}, improving inflammation and oxidative stress^{(Reference Derosa, D’angelo and Maffioli75)}, as well as anti-hypertensive properties^{(Reference Price, Jackson and Lovegrove76)}. WP are extracted from the liquid whey fraction of dairy milk which is a by-product of the cheese making process. With the invention of membrane filtration technology in the 1970s it became possible to produce concentrated whey powders of up to 80 g/100 g protein content from liquid whey, known as whey protein concentrate (WPC). Further processing through micro-filtration produces whey protein isolate (WPI) which contains up to 90 g/100 g protein. WPC can be enzymatically hydrolysed to produce a partially pre-digested product known as whey protein hydrolysate (WPH). This is mainly used in infant formulas for babies who cannot tolerate standard formulas, and has antioxidant properties, making it useful as an additive in processed foods to prevent oxidative degeneration^{(Reference Vavrusova, Pindstrup and Johansen77)}.

WPC and WPI are both used as sports and dietary nutritional protein supplements^{(Reference Guo78)}. The market for WP in the form of supplementary protein shakes, bars and as a protein booster in many other foods has grown exponentially in the last 20 years^{(Reference Keogh, Li and Gao8)}. Protein has become a very fashionable macronutrient, popularly understood to support lean muscle tissue, recovery from exercise, increase satiety and aid in fat loss or body weight maintenance^{(Reference Harwood and Drake79)}. Supplementary WP is now being consumed in amounts far greater than ever before⁽⁸⁰⁾ with the UK market predicted to grow by nearly 5% a year until 2032⁽⁸¹⁾. However, there is limited data available on specific doses and frequency of consumption in the general population in the UK.

WP and markers of IR and glycaemic control

Despite WP being the richest dietary source of BCAA, and the established link between elevated fasting plasma BCAA concentrations and increased risk of T2D, consumption of bovine dairy products, particularly WP, has been associated with a lower risk of developing T2D in large prospective cohort studies^{(Reference Tong, Dong and Wu82)}. Research interest has grown in the potential role of WP as a functional food for managing or preventing T2D and associated risk factors. Human trials in both healthy individuals and those with IR or T2D have shown that WP supplementation can have a favourable effect on glycaemic control biomarkers by enhancing insulin response and lowering postprandial glucose levels^{(Reference Akhavan, Luhovyy and Brown83–Reference King, Walker and Campbell86)}. However, a recent systematic review and meta-analysis suggests the current evidence for a beneficial effect of WP on postprandial glucose levels is ‘very low to low’ certainty, with further research needed^{(Reference Chiang, Liu and Loh74)}. Chronic studies have similarly indicated that WP may improve fasting glycaemia, reduce markers of inflammation and oxidation, and beneficially influence cardiovascular disease risk factors. A comprehensive umbrella review by Connolly et al. ^{(Reference Connolly, Wang and Bergia87)}, which included both chronic and acute studies, supported these findings, showing WP’s potential to lower HbA1c, HOMA-IR and fasting insulin, without any reported adverse effects on T2D-related risk factors.