Milk and dairy products are widely consumed around the world on a daily basis. They are not only an important source of nutrients in the human diet, but they also represent important value in the food chain providing opportunities for farmers, food processors and retailers to contribute to increased food security and poverty alleviation( Reference Muehlhoff and McMahon 1 ). Therefore any change in milk and dairy consumption will have multiple impacts on human and animal health, environment, food security and economics. Indeed, according to an OECD-FAO report, milk production is projected to increase by 180 million tonnes in the next decade, predominantly in developing countries( 2 ). Moreover, the inclusion of animal-derived products adds diversity to plant-based diets, providing an important source of many essential nutrients, the dietary requirements of which would be more difficult to meet by plant-based diets. However, the potential health impacts of animal-derived foods, and more specifically milk and dairy consumption, have been questioned owing to their high saturated fat content (for review, see( Reference Markey, Vasilopoulou and Givens 3 )). However, emerging epidemiological evidence supports the beneficial effects of milk and dairy consumption on health, particularly cardiometabolic health( Reference Elwood, Pickering and Givens 4 – Reference Aune, Norat and Romundstad 6 ).
Milk is a complex food, a unique package of many nutrients such as calcium, magnesium, iodine, phosphorus, vitamin B12, pantothenic acid, riboflavin, high-quality protein, peptides and oligosaccharides. In the human body, these bioactive components may interact with each other and exert synergistic effects, making it difficult to assign the specific health effect of a single component. Bovine milk, which is widely consumed around the world, contains approximately 32–34 g/l protein of which 80 % (w/w) is casein and 20 % (w/w) is whey protein. Both milk proteins consist of smaller protein fractions such as casein: α-s1, α-s2, β and κ-casein; and whey: β-lactoglobulin, α-lactalbumin, lactoferrin, immunoglobulins, serum albumin, glycomacropeptide, enzymes and growth factors. Milk proteins are considered to be high-quality proteins. Whey protein is rich in branched-chain amino acids (AA; BCAA) such as leucine, isoleucine and valine, whilst casein contains more histidine, methionine, phenylalanine, proline, serine, tyrosine and valine. It is well established that casein and whey have differential effects on gastric emptying and kinetics of digestion and absorption( Reference Boirie, Dangin and Gachon 7 ). Intact micellar casein clots in the stomach due to the low pH, and is, therefore, digested more slowly, which results in a prolonged and more sustained AA release. In contrast, intact whey (which is acid soluble) or hydrolysed whey and caseinate are absorbed more rapidly, with a faster AA release and half-life( Reference Boirie, Dangin and Gachon 7 ). It is, however, of note that micellar casein is different from Ca or Na caseinate (micellar casein is acidified and neutralised with alkali e.g. NaOH or Ca(OH)2 in order to form caseinate), as the latter are soluble and thus may show similarities to whey in terms of digestion rates( Reference Phillips 8 , Reference Reitelseder, Agergaard and Doessing 9 ). As a result of their different inherent AA compositions leading to distinct absorption and kinetic behaviour, they may also have differential effects on human health.
The aim of this review is to update and critically evaluate the existing evidence on the effects of casein and whey on metabolic function, including blood pressure (BP), vascular function, glucose and lipid metabolism, and inflammation.
Comprehensive literature search
A comprehensive literature search was conducted using the electronic databases Medline, the Cochrane Library, EMBASE and Web of Science using the following terms: intervention, randomised controlled trials (RCT), clinical trials, high BP, hypertension, anti-hypert*, vascular function, endothelial function, vascular stiffness, milk protein, milk peptide*, casein, hydrolysate, human subjects, lipids, insulin, glucose, inflammation. Furthermore, hand-searching was performed on the reference lists of both studies and review articles. In addition, Google and Google Scholar were used to confirm that the search was complete. The search period covered studies published until September 2015.
Blood pressure
CVD remain the leading cause of death in most countries worldwide. In the UK, there has been a significant decrease in death rates since 1961, and due to a combination of better healthcare and preventative strategies, in 2012 CVD became the second main cause of death (CVD caused 28 % of all death and cancer 29 %)( Reference Townsend, Bhatnagar and Wickramasinghe 10 ). Approximately seven million people live with CVD in the UK, which costs £19 billion each year (including premature death, lost productivity, hospital treatment and prescriptions) resulting in a significant economic burden( Reference Townsend, Bhatnagar and Wickramasinghe 10 ). Premature death from CVD can be prevented by improving modifiable risk factors. For example, it has been estimated that in the general population increasing physical activity, smoking cessation and dietary changes can lead to 50, 20–30 and 15–40 % mortality risk reduction, respectively( Reference Iestra, Kromhout and van der Schouw 11 ).
High BP (hypertension) is the key modifiable risk factor of CVD and of stroke in particular. Nearly 30 % of adults in the UK have high BP; however, only half of them are aware of it and even less receive treatment( Reference Townsend, Bhatnagar and Wickramasinghe 10 ). High BP is present when systolic blood pressure (SBP) is ≥140 mmHg and/or diastolic blood pressure (DBP) is ≥90 mmHg( Reference Mancia, De Backer and Dominiczak 12 ). It is important to treat hypertension and maintain BP in the normal range as elevated BP can cause irreversible damage to different organs such as kidneys, heart and eyes( Reference Mancia, De Backer and Dominiczak 12 ).
Long-term studies on blood pressure
We have recently reviewed the evidence from RCT on the antihypertensive effects of milk proteins and peptides( Reference Fekete, Givens and Lovegrove 13 ). For that review we systematically searched and reviewed the literature until December 2012. There was an imbalance in the literature as more RCT were conducted using mainly one type of casein-derived peptides, called lactotripeptides (LTP). We, therefore conducted an updated meta-analysis on the impact of LTP on BP( Reference Fekete, Givens and Lovegrove 14 ), which included all available and relevant RCT and detailed subgroup and regression analyses, which were somewhat limited in previous meta-analyses in this area( Reference Xu, Qin and Wang 15 – Reference Qin, Xu and Dong 18 ). We found a small, but significant reduction in both SBP (−2·95 mmHg (95 % CI −4·17, −1·73; P < 0·001)) and DBP (−1·51 mmHg (95 % CI −2·21, −0·80; P < 0·001)) after 4 weeks of LTP supplementation in pre- and hypertensive populations. Since there was a statistically significant heterogeneity of treatment effects across studies, sub-group analyses were performed. These analyses suggested differences in countries where RCT were conducted: Japanese studies reported significantly greater BP-lowering effect of LTP (−5·54 mmHg for SBP; and −3·01 mmHg for DBP), compared with European studies (−1·36 mmHg for SBP; and −0·83 mmHg for DBP; P = 0·002 for SBP and <0·001 for DBP). This was confirmed in a recent meta-analysis which focused on Asian RCT only. However, it only assessed SBP and the authors reported a very similar reduction of −5·63 mmHg in SBP compared to our −5.54 mmHg( Reference Chanson-Rolle, Aubin and Braesco 19 ). There may be several explanations for this observation. Firstly Japanese diets contains less milk and dairy products than European diets, therefore consumption of milk proteins may have a greater overall impact when compared with populations that consume these proteins more regularly and in higher quantities( 20 ). Furthermore, there are reported ethnic differences in the response to drug administration, BP lowering in particular( Reference Johnson 21 ) which could impact on the response to these bioactive proteins. Finally differences in response may also have resulted from different spatial conformations (cis/trans) of LTP used in the studies, due to production processes( Reference Siltari, Viitanen and Kukkurainen 22 ). Intriguingly, we also found a small-study effect, and when all bias was considered it shifted the treatment effect towards a less significant SBP and non-significant DBP reduction in response to LTP supplementation. We concluded that with potential bias considered, LTP consumption may still be effective in lowering BP in mildly hypertensive or hypertensive groups( Reference Fekete, Givens and Lovegrove 14 ).
During our systematic literature search( Reference Fekete, Givens and Lovegrove 13 ) we found that there were very few studies investigating the BP-lowering effects of other casein-derived peptides in human subjects( Reference Ashar 23 – Reference Sugai 27 ). Furthermore these studies were limited, used different types of peptides and were often uncontrolled with poor methodological and study design. Due to these inconsistencies in the study design, it was impossible to compare these data and no firm conclusion could be drawn on the antihypertensive effects of casein-derived peptides. Similarly, we found a limited number of RCT conducted using intact whey or whey-derived peptides assessing their antihypertensive effects in human subjects ( Reference Kawase, Hashimoto and Hosoda 28 – Reference Pal and Ellis 33 ). These trials seem to be of higher quality than studies on casein-derived peptides; however, the findings of these studies were also inconsistent( Reference Fekete, Givens and Lovegrove 13 ).
Since our review, published in 2013, three new studies which assessed the effects of milk proteins on BP as primary outcome were published. Petyaev et al. ( Reference Petyaev, Dovgalevsky and Klochkov 34 ) examined the impacts of whey protein embedded in a protective lycopene matrix, a new proprietary formulation, the so-called whey protein lycosome, in a pilot study. Authors hypothesised that this formulation would protect whey protein from gastrointestinal degradation which would increase the bioavailability of the protein, and thus reduce the need for a high dose. They administered 70 mg whey protein along with 7 mg lycopene in the form of a capsule (WPL) and compared this with whey protein (70 mg) and lycopene (7 mg) separately (taken once daily for a month). A significant decrease in BP (−7 mmHg in SBP and −4 mmHg in DBP, P < 0·05) in the WPL group was reported compared with baseline only and no effect relative to the whey and lycopene given separately. Due to the nature of this pilot study, there was no information on blinding, the sample size was small (ten/treatment group) and due to the limited statistical analysis further investigation is needed to evaluate the potential antihypertensive effect of WPL. Another RCT was conducted in overweight and obese adolescents (aged 12–15 years), who were asked to consume 1 litre/d of either water, skimmed milk, whey or casein (milk-based treatment drink contained 35 g/l protein) for 12 weeks( Reference Arnberg, Larnkjaer and Michaelsen 35 ). A decrease in brachial and central aortic DBP compared with baseline and control group (consuming water) was observed, whereas whey protein appeared to increase brachial and central aortic SBP, and central DBP. The authors acknowledged several limitations of the study, including difficulties in recruitment, changes in the research protocol after study commencement and not controlling for the extra energy intake that 1 litre/d treatment drinks provided, which led to an increase in weight in those in the treatment groups compared with a loss in the control group which consumed water. Therefore, due to these limitations it was difficult to draw firm conclusions from these data. A study of Figueroa et al. ( Reference Figueroa, Wong and Kinsey 36 ) examined the effects of both whey and casein on BP and vascular function combined with exercise training in obese, hypertensive women. In their 4-week trial, participants were assigned to consume 30 g casein, whey or 34 g maltodextrin (control) and perform resistance and endurance exercises 3 d/week under a qualified instructor's supervision. They reported significant reduction in both brachial and aortic SBP in both whey and casein groups compared with the control, although this was not observed for DBP. The exercise training did not have additional effects on BP or arterial function, owing the beneficial effect on the cardiovascular system to the milk proteins (Table 1).
Table 1. Impacts of milk proteins on blood pressure
↑, Increase; ↓, decrease; BP, blood pressure; bBP, brachial blood pressure; cBP, central blood pressure; DBP, diastolic blood pressure; SBP, systolic blood pressure.
In summary, emerging evidence suggest that milk protein consumption for at least 4 weeks may result in small BP lowering; however, further well-controlled studies involving 24-h ambulatory BP monitor should be performed for confirmation.
Short-term studies on blood pressure
According to a typical Western eating pattern, people spend up to 18 h/d in a postprandial state consuming three or more meals daily. Furthermore elevated postprandial lipeamia, glycaemia and inflammation have been linked with increased risk for chronic disease development, including diabetes and CVD( Reference Alipour, Elte and van Zaanen 37 – Reference Lopez-Miranda, Williams and Lairon 39 ). Therefore dietary strategies that attenuate the postprandial metabolic disturbance are urgently required.
To date only two studies have evaluated the acute (short-term) effects of milk proteins on BP. Pal and Ellis compared 45 g whey protein isolate, 45 g Na-caseinate with 45 g glucose in conjunction with a breakfast in normotensive overweight and obese women( Reference Pal and Ellis 32 ), but found no effect of treatment. A more recent study compared the postprandial effects of several dietary proteins (milk protein, pea protein and egg-white) and carbohydrate-rich meals on BP-related responses( Reference Teunissen-Beekman, Dopheide and Geleijnse 40 ). Although the authors failed to specify the specific type of milk protein isolate used, its BP-lowering effect was not significantly different to pea protein, although both milk and pea protein were significantly lower than egg-white (P ≤ 0·01; Table 1). The lack of evidence on the acute BP effects of milk proteins warrants further research.
Vascular function
Vascular dysfunction is often used as an umbrella term for abnormalities of the vascular system, such as endothelial dysfunction and arterial stiffness( Reference Schachinger, Britten and Zeiher 41 ). The endothelium, the inner layer of cells of the vasculature, plays a key regulatory role in the vascular system. Any disturbance in endothelial function, such as increased permeability, reduced vasodilation and activation of thrombotic and inflammatory pathways, can lead to atherosclerotic development( Reference Verma and Anderson 42 ). Due to the central role of the endothelium in the development of atherosclerosis, several non-invasive methods have been developed to assess endothelial dysfunction. Flow-mediated dilation (FMD) is considered to be the gold standard method of assessing endothelial function and may surpass the predictive value of traditional risk factors such as smoking, elevated cholesterol level in predicting cardiovascular events in patients with established CVD( Reference Thijssen, Black and Pyke 43 ). However, it is of note that this technique requires extensive training and is operator dependent, which may limit its value.
Arterial stiffness is a measure of arterial elasticity which is the ability to expand and contract along with cardiac pulsation and relaxation. CVD risk factors such as ageing, hypertension, smoking and diet have been shown to have a detrimental effect on arterial distensibility, inducing an imbalance between the synthesis and degradation of elastin and types 1 and 3 collagen( Reference Bruno, Bianchini and Faita 44 ). Pulse wave velocity is considered to be the gold standard to measure arterial stiffness and has a substantial predictive value for CVD events( Reference Vlachopoulos, Aznaouridis and Stefanadis 45 ).
Long-term studies on vascular function
Our previous review also evaluated the health effects of milk proteins and/or their peptides on vascular function( Reference Fekete, Givens and Lovegrove 13 ). In brief, we identified nine chronic RCT( Reference Pal and Ellis 33 , Reference Hirota, Ohki and Kawagishi 46 – Reference Nakamura, Mizutani and Ohki 53 ), of which eight used LTP( Reference Hirota, Ohki and Kawagishi 46 – Reference Turpeinen, Ehlers and Kivimaki 54 ) and one trial used intact casein and whey( Reference Pal and Ellis 33 ). These studies were diverse in several aspects of methodologies such as design, length and dose of treatment, subject characteristics and measures of vascular function, and most importantly type of milk proteins used. Due to this heterogeneity, it is not possible to draw firm conclusions on the relative effects of milk proteins on the vascular function.
We have identified three further RCT: Petyaev et al. ( Reference Petyaev, Dovgalevsky and Klochkov 34 ) examined the impacts of WPL not only on BP, but also on vascular reactivity, using FMD. They reported statistically significant improvements in FMD in the WPL group only (+2·6 %, P < 0·05) compared with baseline. Arnberg et al. also evaluated the effects of intact whey, casein and semi-skimmed milk on arterial stiffness using pulse wave velocity, however, failed to show any changes in vascular function( Reference Arnberg, Larnkjaer and Michaelsen 35 ). Figueroa et al. ( Reference Figueroa, Wong and Kinsey 36 ) reported favourable changes in augmentation index (a measure of arterial stiffness) and brachial-pulse wave velocity in both whey and casein groups combined with exercise, compared with the control group. It is of note that the randomisation may not have been adequate as the baseline values for both BP and arterial stiffness were different in the treatment groups, which may have confounded the study (Table 2).
Table 2. Impacts of milk proteins on vascular function
FMD, flow-mediated dilation, ↑, increase; ↔, no effect.
Short-term studies on vascular function
Only four RCT were conducted to evaluate the effects of milk proteins on vascular function in a postprandial setting( Reference Pal and Ellis 32 , Reference Turpeinen, Ehlers and Kivimaki 54 – Reference Ballard, Bruno and Seip 56 ). Pal and Ellis failed to show any acute effects of whey and casein ingestion with a meal in normotensive obese postmenopausal women on arterial stiffness measured by pulse wave analysis( Reference Pal and Ellis 32 ). Likewise, Turpeinen et al. ( Reference Turpeinen, Ehlers and Kivimaki 54 ) also did not observe any statistically significant change in arterial stiffness measured by pulse wave velocity after acute ingestion of 25 mg LTP with 2 g plant sterol ester mixed in a milk drink in mildly hypertensive subjects. However, Ballard et al. reported significant improvements in arterial reactivity assessed by FMD (+4·3 %) at 120 min after ingestion compared with placebo corresponding time point (P < 0·05) in mildly hypertensive, overweight individuals after whey hydrolysate (5 g NOP-47) ingestion with water( Reference Ballard, Kupchak and Volk 55 ). Mariotti et al. failed to report any significant effects of casein, whey or α-lactalbumin enriched whey protein on digital volume pulse (a measure of arterial stiffness)( Reference Mariotti, Valette and Lopez 57 ) (Table 2).
Intriguingly, BP-lowering effects of milk proteins were not associated with changes in vascular function in the reviewed RCT( Reference Fekete, Givens and Lovegrove 13 ) which is confirmed by emerging evidence on the relationship between BP and arterial stiffness. This suggests that the interaction between BP and arterial stiffness may be bi-directional( Reference Franklin 58 , Reference Dernellis and Panaretou 59 ) via complex interactions between different pathways such as inflammatory( Reference Sesso, Buring and Rifai 60 , Reference Tsai, Lin and Lin 61 ), hormonal (e.g. leptin and insulin)( Reference Tsai, Lin and Lin 61 – Reference Singhal, Farooqi and Cole 63 ) and disturbances in endothelial-derived mediators( Reference Franklin 58 ). Therefore it is important to determine the effects of milk proteins on other mediators of CVD risk that may indirectly affect BP.
Glycaemic control
Insulin has a range of biological actions within the human body( Reference Baron 64 ), it not only has a key regulatory role in metabolic energy disposal and storage in tissues, but also it is responsible for cell growth and development( Reference Rosen 65 ), ion transport( Reference Moore 66 ) and sympathetic nervous system activity( Reference Anderson, Hoffman and Balon 67 ). In addition, insulin has haemodynamic activities such as increasing blood flow and cardiac output, probably via increased NO production( Reference Baron 64 ). Giugliano et al. demonstrated insulin release after an intravenous infusion of l-arginine resulted in improvements in FMD( Reference Giugliano, Marfella and Verrazzo 68 ). However, Gates et al. ( Reference Gates, Boucher and Silver 69 ) showed an insulin-independent vasodilation after l-arginine administration. Similarly, Ballard et al. reported an insulin-independent FMD improvement in response to the acute ingestion of a whey-derived peptide, NOP-47( Reference Ballard, Kupchak and Volk 55 ).
It is well established that food proteins and more specifically AA acutely stimulate insulin secretion( Reference Floyd, Fajans and Conn 70 ) with several AA possessing direct insulinotropic effects( Reference Schmid, Schusdziarra and Schulte-Frohlinde 71 , Reference Schmid, Schulte-Frohlinde and Schusdziarra 72 ). Both whey and casein appear to increase insulin secretion, however, to different extents( Reference Nilsson, Stenberg and Frid 73 ). This may be due to their effect on gastric emptying, absorption and kinetics, since the insulin responses seemed to correlate with the increase in plasma AA concentration after protein ingestion( Reference Calbet and MacLean 74 ). Likewise, hydrolysates appear to increase insulin production more than intact proteins( Reference Calbet and Holst 75 ).
It is not yet known how milk proteins exert their beneficial effects on glucose homeostasis; however, BCAA, in particular, leucine, isoleucine, valine, lysine and threonine are shown to act as insulin secretagogues (inducing insulin secretion from pancreatic β-cells), with leucine reportedly having the greatest insulinotopic effect acutely( Reference van Loon, Saris and Verhagen 76 ). This may be via the regulation of both ATP production (by metabolic oxidation and allosteric activation of glutamate dehydrogenase) and K ATP activity( Reference Yang, Chi and Burkhardt 77 ). Similarly, BCAA and particularly leucine, have been reported to activate the mammalian rapamycin pathway resulting in a higher incretin hormone (insulin, glucagon-like peptide 1 (GLP-1) and gastric inhibitory peptide (GIP)) synthesis( Reference Yang, Chi and Burkhardt 77 , Reference Melnik 78 ). GIP is also known as glucose-dependent insulinotropic peptide, synthesised by K cells found in the mucosa of the duodenum and jejunum in response to food ingestion, which may subsequently further induce insulin production( Reference Yabe and Seino 79 ). While the effect of GIP appears to be more pronounced at normoglycaemic levels, GLP-1 is more active during hyperglycaemia( Reference Yabe and Seino 79 ). Jakubowicz and Froy showed that whey protein drink increased GIP response (+80 %) in healthy adults, yet a mixture of BCAA mimicking the supply of AA in whey protein, failed to exert the same effect( Reference Jakubowicz and Froy 80 ). Therefore they suggested that certain bioactive peptides and/or AA deriving from whey protein during digestion may be responsible for this action( Reference Jakubowicz and Froy 80 ). GLP-1 is a potent antihyperglycaemic hormone secreted by intestinal L cells( Reference Yabe and Seino 79 ). Interestingly, it has been shown to possess cardioprotective effects, which may be further complemented by natriuretic and antioxidative stress on the kidneys leading to beneficial impacts on BP and vasculature( Reference Poudyal 81 ). This warrants further consideration in future research when the effects of milk proteins on the cardiovascular system are assessed. Additionally, GLP-1 was more pronounced in healthy subjects after whey consumption compared with casein or soya; however, after 2 h of ingestion the concertation of the hormone decreased, while it continued to increase after casein( Reference Jakubowicz and Froy 80 , Reference Hall, Millward and Long 82 , Reference Veldhorst, Nieuwenhuizen and Hochstenbach-Waelen 83 ). This may be explained by the different plasma kinetics of milk proteins. Two enzyme inhibitory peptides deriving from milk proteins have been associated with the beneficial effects on the glucose homeostasis: dipeptidyl peptidase-IV enzyme inhibitors and α-glucosidase enzyme inhibitors. Although dipeptidyl peptidase-IV plays several roles in different physiological processes, it has a distinct effect on glucose homeostasis by degrading incretin hormones GLP-1 and GIP( Reference Fan, Yan and Stehling 84 ). Whereas there is a definite lack of human studies examining the effects of dipeptidyl peptidase-IV inhibitory peptides deriving from milk proteins; some in silico (computer-aided), in vitro and limited animal studies suggest a potential role in controlling glucose metabolism. Lacroix and Li-Chan proposed that casein appears to be a better source of dipeptidyl peptidase-IV inhibitory peptides than whey protein( Reference Lacroix and Li-Chan 85 ). However, in vitro and in vivo studies suggest that whey protein may be equal or a better source of these inhibitory peptides (for review see( Reference Patil, Mandal and Tomar 86 )). The α-glucosidase enzyme is found in the brush border of the enterocytes in the small intestine and is responsible for the synthesis and breakdown of carbohydrate by cleaving glycosidic bonds in complex carbohydrates to produce monosaccharides. A potential therapy in type 2 diabetic patients could be to reduce the absorption of glucose by carbohydrate hydrolysing enzymes such as α-glucosidase, which may also enhance and promote GLP-1 secretion( Reference Slama, Elgrably and Sola 87 ). A very limited number of in vitro studies demonstrated that α-glucosidase inhibitory peptides may be derived from whey protein( Reference Lacroix and Li-Chan 88 , Reference Konrad, Anna and Marek 89 ). This clearly warrants further research.
Short-term studies on glycaemic control
Milk proteins have been extensively investigated for their insulinotropic and glucose-lowering effects in healthy subjects( Reference Nilsson, Stenberg and Frid 73 , Reference Calbet and Holst 75 , Reference Hall, Millward and Long 82 , Reference Veldhorst, Nieuwenhuizen and Hochstenbach-Waelen 83 , Reference Petersen, Ward and Bastian 90 – Reference Holmer-Jensen, Hartvigsen and Mortensen 99 ) and to a limited extent in individuals with suboptimal glucose control( Reference Frid, Nilsson and Holst 100 – Reference Geerts, van Dongen and Flameling 106 ). The dose varied significantly between studies from as little as 10 g( Reference Akhavan, Luhovyy and Panahi 92 , Reference Jonker, Wijngaarden and Kloek 105 , Reference Geerts, van Dongen and Flameling 106 )–51 g( Reference Pal and Ellis 91 ). Milk proteins were administered on their own or with a meal or even served as pre-meals. Current evidence on the effects of whey protein on glucose control appears to be more promising than casein; furthermore it has been proposed that whey protein may be as effective at inducing insulin secretion as medication (sulfonylureas) prescribed for management of hyperglycaemia in type 2 diabetic patients( Reference Jakubowicz and Froy 80 , Reference McGregor and Poppitt 107 ) (Table 3). Thus, providing a rationale for individuals with impaired glucose control or for patients with type 2 diabetes mellitus to consume whey protein prior to or with meals to control postprandial glucose metabolism. Future studies should examine the minimum dose at which whey protein exerts beneficial effects. Similarly due to the different time-frame by which milk proteins have an effect, longer postprandial trials (e.g. 24 h) may provide important information on how casein could improve hyperglycaemia in individuals characterised by insulin resistance but with functional β-cells.
Table 3. Impacts of milk proteins on glycaemic control
↑, Increase; ↓, decrease; ↔, no effect; CH, casein hydrolysate; D, day; GIP, glucose-dependent insulinotropic polypeptide; GLP-1, glucagon-like peptide 1; HC; hydrolysed casein; HOMA-IR, homeostasis model assessment of insulin resistance; MS; metabolic syndrome; T2D, type-2 diebetes; Whey MPM, whey malleable protein matrix; WP, whey protein; WPH, whey protein hydrolysate; WPI, whey protein isolate.
Long-term studies on glycaemic control
To the best of our knowledge, only three studies have investigated the chronic supplementation of milk proteins, rather than milk or dairy products, on glycaemic control. Pal et al. examined the effects of whey and casein (2 × 27 g/d for 12 weeks) in overweight and obese subjects( Reference Pal, Ellis and Dhaliwal 96 ). Most subjects had borderline impaired glucose tolerance at baseline, but at the end of the intervention a reduced fasting insulin concentration was observed in the whey protein group compared with the control group (glucose), although no change in fasting glucose was reported. In another study, a whey fermentation product (malleable protein matrix) decreased fasting plasma glucose concentration after 3 months supplementation compared with the control group, which was more pronounced in individuals with impaired fasting glucose at baseline( Reference Gouni-Berthold, Schulte and Krone 108 ). An acute-in-chronic study also reported a decrease in postprandial glucose response in whey group, which remained unchanged after the 4-week supplementation period( Reference Ma, Jesudason and Stevens 102 ) (Table 3).
Lipid metabolism
Short-term studies on lipids
Postprandial triacylglycerolaemia has been associated with markers of early atherosclerosis such as endothelial dysfunction and carotid media thickness( Reference Teno, Uto and Nagashima 109 , Reference Vigna, Delli Gatti and Fellin 110 ) and is strongly influenced by the composition of a meal, including the quality and quantity of fat( Reference Thomsen, Storm and Holst 111 , Reference Thomsen, Rasmussen and Lousen 112 ) and carbohydrate( Reference Cohen and Schall 113 , Reference Lairon, Play and Jourdheuil-Rahmani 114 ). In theory due to the insulinogenic effects of milk proteins, their consumption would be predicted to attenuate postprandial lipaemia, as insulin has an inhibitory effect on hormone-sensitive lipase and hepatic release of free fatty acid and stimulatory effect on lipoprotein lipase which hydrolyses TAG for metabolism or storage. However, evidence from postprandial RCT is limited. Postprandial investigations reported decrease in TAG after both whey and casein ingestion in combination with a fat-rich meal in obese( Reference Holmer-Jensen, Mortensen and Astrup 98 ) and individuals with type 2 diabetes mellitus ( Reference Mortensen, Hartvigsen and Brader 103 , Reference Brader, Holm and Mortensen 115 ), but showed no effect on TAG after acute consumption of whey protein( Reference Holmer-Jensen, Hartvigsen and Mortensen 99 , Reference Mortensen, Holmer-Jensen and Hartvigsen 104 ). Free fatty acid also decreased after whey and casein ingestion in obese( Reference Holmer-Jensen, Hartvigsen and Mortensen 99 ) and type 2 diabetes mellitus patients( Reference Mortensen, Holmer-Jensen and Hartvigsen 104 ). It is of note that parameters of lipid metabolism such as LDL and HDL and total cholesterol remain stable acutely( Reference Olefsky, Crapo and Reaven 116 , Reference Roche, Zampelas and Knapper 117 ).
Recently an acute study reported that casein with a high-fat, high-energy meal, compared with whey protein and α-lactalbumin-enriched whey protein, significantly reduced postprandial TAG and had a marked effect of chylomicron kinetics( Reference Mariotti, Valette and Lopez 57 ). This could be due to the different physicochemical makeup of casein and whey protein, as casein forms a gel in the stomach influencing the rate of absorption and gastric emptying (Table 4).
Table 4. Impacts of milk proteins on lipid metabolism
↑, Increase; ↓, decrease; ↔, no effect; CH, casein hydrolysate; D, day; FFA, free fatty acids; HC; hydrolysed casein; HDL-c, HDL cholesterol; LDL-c, LDL cholesterol; MS; metabolic syndrome; T2D, type-2 diebetes; TC, total cholesterol; Whey MPM, whey malleable protein matrix; WP, whey protein; WPH, whey protein hydrolysate;WPI, whey protein isolate.
Long-term studies on lipids
To date, five chronic RCT, which examined the lipid-lowering effects of milk proteins, have been identified. Three month supplementation of whey (2 × 25 g/d) and casein (2 × 25 g/d) during an ad libitum weight regain diet after substantial diet-induced weight loss in healthy obese subjects resulted in no change in plasma lipids( Reference Claessens, van Baak and Monsheimer 118 ). However, whey protein isolate (2 × 27 g/d) significantly reduced fasting TAG, total cholesterol and LDL-cholesterol after 3 months in overweight, obese individuals( Reference Pal, Ellis and Dhaliwal 96 ). Another 3-month supplementation study with malleable protein matrix (15 g/d protein in two daily servings of 150 g yoghurt) reduced fasting TAG, which was more pronounced in subjects with elevated baseline TAG( Reference Gouni-Berthold, Schulte and Krone 108 ). In a 6-week study casein (35 g/d) also reduced total cholesterol in hypercholesterolaemic subjects( Reference Weisse, Brandsch and Zernsdorf 119 ). Petyaev et al. reported a decrease in LDL-cholesterol, TAG and total cholesterol in their pilot study( Reference Petyaev, Dovgalevsky and Klochkov 34 ) (Table 4). The limited evidence suggests that milk proteins have a beneficial impact on fasted lipids; however further studies are required. Although its possible mechanism of action is not clear, insulin may play a role. In vitro studies suggest that milk proteins and BCAA inhibit expression of genes involved in intestinal fatty acid and cholesterol absorption and synthesis( Reference Chen and Reimer 120 ). Whey has been shown to induce urinary excretion of tricarboxylic acid cycle compounds such as citric acid and succinic acid in rats, which are substrates for lipogenesis, suggesting an increased catabolic state (e.g. lipolysis) and reduced lipid accretion compared with casein( Reference Lillefosse, Clausen and Yde 121 ). This could be a possible mechanism of lipid reduction. Similarly, in another metabolic study conducted in human subjects, cheese (casein) appeared to induce lowering of urinary citrate( Reference Zheng, Yde and Clausen 122 ), which suggests that cheese consumption affects the tricarboxylic acid cycle. Additionally, microbiota-related metabolite, hippuric acid was significantly higher in the cheese group, than in the milk, implying a stimulation of gut bacteria activity. The enhanced bacterial activity also resulted in higher SCFA( Reference Zheng, Yde and Clausen 122 ), which have been proposed as key regulatory metabolites in lipid metabolism( Reference Tremaroli and Backhed 123 ). This effect may be due to the cheese matrix rather than the casein per se. An in vivo study proposed another potential mechanism of action through decreased lipid infiltration into the liver in rats with non-alcoholic fatty liver( Reference Hamad, Taha and Abou Dawood 124 ). Another possible putative mechanism is increased fat oxidation. Lorenzen et al. ( Reference Lorenzen, Frederiksen and Hoppe 125 ) demonstrated an increased lipid oxidation after acute casein consumption compared with whey. They speculated that it may be due to lower insulin secretion after casein consumption relative to whey since insulin down-regulates lipid oxidation. However, insulin was not measured in the study and this mechanism could not be confirmed. The same research group examined the effects of dairy Ca on lipid metabolism in conjunction with a low- and high-fat diet for 10 d( Reference Lorenzen and Astrup 126 ). They found that dairy Ca attenuated the increase in total and LDL-cholesterol, without affecting the rise in HDL-cholesterol. This observed phenomenon may be due to the formation of insoluble Ca-fatty acid soaps and/or the production of hydrophobe aggregation with bile and with other fatty acids( Reference Lorenzen and Astrup 126 – Reference Govers, Termont and Van Aken 128 ).
Inflammation and oxidative stress
Inflammation and oxidative stress are chronic conditions which contribute to many diseases such as obesity( Reference Biro and Wien 129 ), type 2 diabetes mellitus ( Reference Conen, Rexrode and Creager 130 ) and CVD( Reference Libby, Ridker and Hansson 131 ). Different dietary components have an impact on low-grade inflammation( Reference Galland 132 ); however, there is a lack of RCT evaluating the acute and chronic consumption of milk proteins on inflammation or oxidative stress with inconsistent outcomes.
Long-term studies on inflammation and oxidative stress
A recent meta-analysis evaluated the effects of chronic consumption of whey protein and hydrolysate on C-reactive protein (CRP), a systemic inflammatory marker( Reference Zhou, Xu and Rao 133 ). Nine RCT were included which showed a small, non-significant reduction in CRP 0·42 mg/l (95 % CI −0·96, 0·13). Sub-group analyses suggested that >20 g/d may be more effective, and the elevated baseline CRP level (≥3 mg/l) could be more responsive to whey or whey peptides consumption( Reference Zhou, Xu and Rao 133 ). Similarly, Arnberg et al. ( Reference Arnberg, Larnkjaer and Michaelsen 35 ) reported no change in CRP in adolescence after whey, casein or skimmed milk consumption for 12 weeks.
IL-6, IL-8 and TNF-α are also recognised inflammatory markers, which induce CRP. Pal and Ellis failed to observe significant changes in these inflammatory markers (2 × 27 g whey or casein or glucose for 12 weeks) in overweight individuals( Reference Pal and Ellis 33 ). However, Sugawara et al. ( Reference Sugawara, Takahashi and Kashiwagura 134 ) reported decreased level of IL-6, IL-8 and TNF-α in patients with chronic obstructive pulmonary disease after whey intervention compared with the control group. Likewise, IL-6 and TNF-α were decreased after lactoferrin consumption for 6 months in postmenopausal women( Reference Bharadwaj, Naidu and Betageri 135 ). Similarly Hirota et al. ( Reference Hirota, Ohki and Kawagishi 46 ) reported decreased levels of TNF-α in mildly hypertensive subjects fed with the casein-derived LTP (Table 5).
Table 5. Impacts of milk proteins on inflammation and oxidative stress
↑, Increase; ↓, decrease; ↔, no effect; BW, body weight; CCL5, CC chemokine ligand-5; CH, casein hydrolysate; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; IPP, isoleucine–proline–proline; MCP-1, monocyte chemotactic protein-1; VPP, valine–proline–proine; Whey MPM, whey malleable protein matrix; WP, whey protein; WPH, whey protein hydrolysate;WPI, whey protein isolate.
Short-term studies on inflammation and oxidative stress
Pal and Ellis also reported no change in IL-6, IL-8 and TNF-α in a postprandial study investigating whey and casein( Reference Pal and Ellis 32 ). Likewise, a whey-derived peptide, NOP-47, also failed to change the level of serum cytokines (TNF-α, IK-6, IL-8, monocyte chemoattractant protein-1, vascular endothelial growth factor, soluble E-selectin, soluble vascular cell adhesion molecule-1) and chemokines( Reference Ballard, Bruno and Seip 56 ). However, consumption of a cake containing whey protein after exhaustive cycling in nine subjects reported reduced levels of CRP and IL-6 by 46 and 50 %, respectively( Reference Kerasioti, Stagos and Jamurtas 136 ). Holmer-Jensen et al. assessed the postprandial effects of whey protein, casein, gluten and cod on low-grade inflammatory markers (monocyte chemotactic protein-1, CC chemokine ligand-5/RANTES (Regulated on activation, normal T cell expressed and secreted)) in conjunction with a high fat meal( Reference Holmer-Jensen, Karhu and Mortensen 137 ). They reported that all meals increased CC chemokine ligand/RANTES; however, the smallest increase was observed after the whey protein meal. Monocyte chemotactic protein-1 was initially suppressed after all meals, and the meal containing whey protein induced the smallest overall postprandial suppression( Reference Holmer-Jensen, Karhu and Mortensen 137 ) (Table 5).
The mechanism of action of milk proteins on oxidative stress and inflammation are unclear but Ca may supresses the pro-inflammatory and reactive oxygen species production in vitro ( Reference Sun and Zemel 138 ). Interestingly, the milk protein-derived inhibitors of the angiotensin-I-converting enzyme may also be involved in the anti-inflammatory process( Reference Kalupahana and Moustaid-Moussa 139 ).
Conclusion and implication for future studies
Taken together, there is a growing number of RCT which suggest that casein and whey protein may have a role in cardiometabolic health. Studies focused on reducing chronic disease risk factors such as hypertension and dysregulated lipid/glucose metabolism by non-pharmacological, dietary strategies will have significant implications not only for social and economic welfare, but also for the healthcare system.
Due to the different physicochemical makeup of casein and whey protein, they may exert differential effects in human subjects. Notably, manufacturing may play a significant role in the physiological effects of milk proteins; however, future studies should investigate which processing method results in more bioactive effects. There is inconclusive evidence on the relative impacts of milk proteins on diurnal BP and vascular function, yet there appears to be strong evidence on the insulinotropic impacts of dairy proteins, owing to the specific AA composition such as BCAA. They also appear to play a beneficial role in lipid homeostasis. Nevertheless the mechanism underlying the action of dairy proteins on the cardiometabolic health warrants further research.
The incorporation of a meal enriched with protein in the habitual diet may result in the improvement of cardiometabolic health as well as the prevention of developing cardiometabolic diseases. Additionally, in contrast with pharmacological antihypertensive treatments, food-derived proteins have not been shown to cause any side-effects or hypotension, making them safe to consume by individuals with a variety of other disease conditions. After careful consideration of the available evidence and knowledge gaps, we have conducted two double-blind, controlled, cross-over studies (Whey2Go studies) aiming to compare the chronic (n 38) and postprandial (n 27) impacts of whey protein (2 × 28 g) and Ca-caseinate (2 × 28 g) with control (2 × 27 g, maltodextrin) on vascular function, BP, markers of insulin resistance, lipid metabolism and inflammatory status in men and women with mild hypertension (≥120/80 mmHg). These studies aim to provide valuable information on the relative effects of milk proteins on BP and on detailed aspects of vascular function compared with maltodextrin. These trials will further our knowledge of whether milk proteins have significant influences as health-promoting food components and whether the public as well as the food industry could benefit. The results from these studies are likely to be available in mid-2016.
Financial Support
This review received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. A. A. F. is supported by a UK Biotechnology and Biological Sciences Research Council (BBSRC) studentship. The human trials were financially supported by BBSRC and Volac Int. Ltd.
Conflicts of Interest
J. A. L. and D. I. G. have previously received funding for research from AHDB Dairy. J. A. L. and D. I. G. have acted as advisors to the Dairy Council. J. A. L. and D. I. G. have received ‘in kind’ foods from Arla for an MRC funded study.
Authorship
A. A. F. conceived and wrote the manuscript. All authors critically reviewed and approved the final version of the manuscript.
