Obesity and body composition

Obesity is one of the strongest risk factors for cardiometabolic disease^{(Reference Field, Coakley and Must9,Reference Rosengren10)} . The increasing global prevalence of obesity has led to a concurrent increase in cardiometabolic disease morbidity and mortality^{(Reference Dai, Alsalhe and Chalghaf11)}. Whilst obesity is clearly associated with cardiometabolic disease, it is a heterogenous condition, with some obese individuals preserving normal cardiometabolic function^{(Reference Blüher12)}. Body composition (or the distribution of adipose tissue) may explain this anomaly. Subcutaneous adipose tissue (SAT), located immediately below the skin, is considered the more ‘metabolically safe’ depot for excess fatty acid storage, whereas visceral adipose tissue (VAT) is located centrally and surrounds intra-abdominal organs^{(Reference Mittal13)}. The latter has been found to be strongly associated with the development of cardiometabolic disease, independently of BMI^{(Reference Hanley, Wagenknecht and Norris14,Reference Smith, Borel and Nazare15)} . The mechanisms driving VAT accumulation, and its consequences on cardiometabolic disease pathophysiology remain an active area of research.

Fatty acid metabolism, inflammation and ectopic lipid deposition

Whilst fasting TAG is associated with both type 2 diabetes and atherosclerotic CVD^{(Reference Zhao, Zhang and Wei16,Reference Sarwar, Danesh and Eiriksdottir17)} , non-fasting TAG has been found to be a more powerful determinant of cardiometabolic risk^{(Reference Nordestgaard, Benn and Schnohr18,Reference Bansal, Buring and Rifai19)} . During fasting conditions, endogenously produced TAG is predominantly transported in hepatically derived VLDL^{(Reference Shelness and Sellers20)}. Postprandially, exogenous (meal-derived) TAG is transported in chylomicrons alongside VLDL–TAG^{(Reference Xiao and Lewis21)}; the majority of TAG is transported to metabolically active tissues in these particles, which are collectively termed TAG-rich lipoproteins^{(Reference Boren, Chapman and Krauss22)}. Upon arrival, TAG-rich lipoproteins are hydrolysed by lipoprotein lipase (LPL), liberating NEFA to be taken up by the tissue. However, a proportion of liberated NEFA escapes to the systemic circulation, termed NEFA spillover^{(Reference Piché, Parry and Karpe23,Reference Barrows, Timlin and Parks24)} . The hydrolysis of TAG-rich lipoproteins leads to the formation of smaller, TAG-poor remnants, which are removed from the circulation by the liver. Importantly, these remnants (particularly chylomicron remnants) may contain a substantial amount of TAG, providing an additional source of TAG for the liver^{(Reference Hultin, Savonen and Olivecrona25)}. An overview of TAG-rich lipoprotein metabolism is shown in Fig. 1. Adipose tissue lipolysis provides further NEFA to metabolically active tissues, and rates of intracellular lipolysis are highest whilst fasting^{(Reference Nielsen, Jessen and Jørgensen26)}. In the transition to the postprandial period, where TAG concentrations are high, adipose tissue lipolysis is suppressed by insulin^{(Reference Zhao, Wu and Rong27)}. In lean/healthy individuals, SAT is responsible for sequestering the majority of fatty acids in the postprandial period. These are subsequently released during fasting, to provide an energy substrate for metabolically active tissues. However, in obesity a number of defects are evident^{(Reference Lewis, Carpentier and Adeli28)}.

Fig. 1.

Metabolism of TAG-rich lipoproteins during the postprandial period. Following the consumption of dietary fat, intestinal enterocytes package meal-derived fatty acids in chylomicrons as TAG. The secreted chylomicron–TAG is then hydrolysed by lipoprotein lipase (LPL) at peripheral tissues, liberating NEFA for uptake. This results in smaller, TAG-poor, chylomicron remnants, which are cleared by the liver along with the remaining TAG in these particles. The liver continues to secrete VLDL–TAG, which is also hydrolysed by LPL at peripheral tissues. This forms intermediate-density lipoprotein (IDL), which is also hydrolysed by LPL and by hepatic lipase (HL), forming TAG-poor low-density lipoproteins (LDL). IDL and LDL are removed from the circulation, predominantly by the liver. Adapted from Borén et al.^{(Reference Borén, Taskinen and Adiels113)}.

During periods of overnutrition, SAT-adipocytes undergo hypertrophic expansion to accommodate excess fatty acids^{(Reference Weyer, Foley and Bogardus29,Reference Verboven, Wouters and Gaens30)} . Increasing adipocyte size may promote hypoxia, apoptosis and endoplasmic reticulum stress, driving tissue damage and promoting macrophage infiltration^{(Reference Weisberg, McCann and Desai31–Reference Frayn and Karpe33)}. Immune cell infiltration, together with large dysfunctional adipocytes, drive the secretion of pro-inflammatory cytokines including: TNF-α, IL-6 and IL-1β, and reduce the secretion of anti-inflammatory cytokines such as IL-10 and IL-5^{(Reference Khanna, Khanna and Khanna34)}. The transition of adipose tissue to a more pro-inflammatory phenotype is thought to be a major driver of chronic low-grade inflammation, which characterises obesity^{(Reference Khanna, Khanna and Khanna34)}.

Hypertrophic SAT-adipocytes also exhibit higher basal rates of lipolysis and have a reduced capacity to store fatty acids^{(Reference Verboven, Wouters and Gaens30,Reference Rutkowski, Stern and Scherer35,Reference Goossens36)} . In the spillover theory^{(Reference Lewis, Carpentier and Adeli28)}, these defects have been hypothesised to redistribute excess fatty acids, driving the accumulation of VAT and ectopic lipid (referring to lipid stored in non-adipose tissues, predominantly the liver, muscle and pancreas). Defective adipose tissue lipolysis, leading to elevated systemic NEFA, was traditionally proposed to be a major driver of this redistribution. However, obesity gives rise to only modest elevations in NEFA; in a systematic review, Karpe et al. reported just a 70 μmol/l difference in NEFA between obese and lean individuals (545 v. 472 μmol/l, respectively), which were unrelated to fat mass^{(Reference Karpe, Dickmann and Frayn37)}. These findings suggest a down-regulation of lipolysis, per weight of adipose tissue, in obesity. Additionally, postprandial meal-derived NEFA spillover has been reported to be lower in obese compared to lean participants, suggesting meal-derived NEFA spillover may in fact be a signature of metabolic health^{(Reference Piché, Parry and Karpe23)}. These findings question the importance of adipose and meal-derived NEFA spillover in the accumulation of VAT and ectopic lipid. Obesity leads to more pronounced increases in fasting and postprandial plasma TAG^{(Reference Lewis, O'meara and Soltys38–Reference Guerci, Verges and Durlach40)}, resulting from an overproduction and reduced clearance of VLDL–TAG and chylomicron–TAG^{(Reference McQuaid, Hodson and Neville39,Reference Shojaee-Moradie, Ma and Lou41,Reference Potts, Coppack and Fisher42)} . Therefore, elevations in TAG and a reduction in SAT clearance of TAG may be a more important driver of VAT and ectopic lipid accumulation.

VAT has been proposed to accelerate ectopic lipid deposition, particularly intrahepatic lipid (IHL)^{(Reference Lewis, Carpentier and Adeli28)}. VAT-adipocytes exhibit increased rates of lipolysis and a more pro-inflammatory profile compared to SAT^{(Reference Verboven, Wouters and Gaens30)}; VAT also drains directly to the liver via the portal circulation, exposing the liver to high concentrations of NEFA and inflammatory markers. Due to these characteristics, elevated VAT is hypothesised to drive IHL accumulation – the ‘portal theory’^{(Reference Lewis, Carpentier and Adeli28)}. Whilst VAT is significantly associated with IHL^{(Reference Thamer, Machann and Haap43)}, the majority of NEFA exposed to the liver are of SAT origin^{(Reference Jensen44)}. In the postprandial period, hepatic uptake of meal-derived NEFA and remnant-TAG may also contribute to IHL deposition^{(Reference Hodson and Frayn45)}. Therefore, it is likely that IHL accumulation results from a number of defective fatty acid metabolism pathways. The accumulation of IHL is proposed to be a primary defect in cardiometabolic disease pathophysiology, and has been shown to be a better marker of the metabolic defects associated with obesity than VAT^{(Reference Fabbrini, Magkos and Mohammed46)}. IHL is associated with a host of metabolic derangements^{(Reference Gastaldelli, Cusi and Pettiti47)} and IHL is also positively associated with VLDL–TAG output^{(Reference Adiels, Taskinen and Packard48)}, which is elevated in obesity^{(Reference Shojaee-Moradie, Ma and Lou41)}. In the twin-cycle hypothesis, elevations in IHL are proposed to drive VLDL–TAG output, leading to excess TAG delivery to the pancreas, thereby increasing intrapancreatic lipid (IPL) accumulation^{(Reference Taylor49)}. Increased VLDL–TAG may also contribute to fatty acid delivery to the muscle, increasing intramyocellular lipid (IMCL)^{(Reference van der Kolk, Goossens and Jocken50)}.

It is apparent that obesity is associated with a plethora of fatty acid metabolism defects. These defects are believed to drive the accumulation of ectopic lipid (primarily IHL, IMCL and IPL), which may promote insulin resistance by lipotoxicity^{(Reference Schaffer51)}. The manifestation of insulin resistance is tissue specific: in the liver it promotes endogenous glucose and VLDL–TAG production, and reduces hepatic insulin clearance^{(Reference Gastaldelli, Cusi and Pettiti47)}; in the pancreas, it drives β-cell dysfunction^{(Reference Tushuizen, Bunck and Pouwels52)}; and in skeletal muscle, it promotes peripheral insulin resistance^{(Reference Anderwald, Bernroider and Krssák53)}. Such defects drive hyperglycaemia and ultimately the development of type 2 diabetes^{(Reference DeFronzo54)}. Furthermore, elevated plasma TAG is associated with the formation of a pro-atherogenic phenotype, whereby excess cholesterol ester transfer protein activity drives a reduction in HDL and elevations in small-dense LDL^{(Reference Lamarche, St-Pierre and Ruel55,Reference Griffin, Freeman and Tait56)} , the latter being an independent predictor of CVD^{(Reference Karalis57)}. These defects are compounded by a state of chronic low-grade inflammation which is typically observed in obesity, driven by elevated VAT and dysfunctional SAT^{(Reference Verboven, Wouters and Gaens30,Reference Karczewski, Śledzińska and Baturo58)} .

The complement system

The complement system was first identified as part of the innate immune system^{(Reference Sim, Schwaeble and Fujita59)}; however, it is now recognised to be a complex network of over fifty plasma and membrane-bound proteins with wide ranging roles in immune, inflammatory and metabolic function^{(Reference Shim, Begum and Yang60)}. As such, markers of the complement system and its activation have been associated with obesity, insulin resistance, type 2 diabetes and CVD^{(Reference Gabrielsson, Johansson and Lönn61–Reference Wlazlo, Van Greevenbroek and Ferreira63)}. Complement activation may result from three major pathways: the classical pathway; mannose-binding lectin pathway and/or the alternative pathway^{(Reference Regal, Gilbert and Burwick64)}. The latter is continuously active at low levels in order to prime complement for rapid activation and may also act as an amplification loop^{(Reference Noris and Remuzzi65)}. All three activation pathways converge on complement component 3 (C3), and its cleavage may ultimately activate the terminal cascade leading to membrane attack complex formation, which can directly lyse pathogens^{(Reference Noris and Remuzzi65)}. Complement activation is controlled by a number of positive and negative regulators^{(Reference Zipfel and Skerka66)}. Upon complement activation, a number of biologically active cleavage products are produced. These include opsonins which covalently bind to target cells (iC3b, C3b and C3d), and anaphylatoxins which are potent pro-inflammatory mediators (C3a and C5a)^{(Reference Noris and Remuzzi65)}. An overview of the complement system is presented in Fig. 2.

Fig. 2.

Complement system. Complement may be activated through three activation pathways: the classical pathway; mannose-binding lectin pathway; and the alternative pathway. All three pathways converge on complement component 3 (C3), and its cleavage may result in terminal pathway activation and the formation of the membrane attack complex (C5b-9n). Complement activation also leads to the production of opsonins and anaphylatoxins (C3a and C5a). Adapted from Regal et al.^{(Reference Regal, Gilbert and Burwick64)}. C1inh, C1 inhibitor; C4BP, C4 binding protein; CD, complement decay accelerating factor; LPS, lipopolysaccharide; MASP, mannose binding lectin associated serine proteases; MBL, mannose binding lectin; MCP, membrane cofactor protein; sC5b-9, inactive membrane attack complex.

The majority of complement components are derived from the liver^{(Reference Gabrielsson, Johansson and Lönn61)}. However, a number of alternative pathway components and regulators are expressed in adipose tissue, including C3, factor B, properdin, factor H and factor I. Furthermore, SAT is the primary site of factor D production^{(Reference Hertle, Stehouwer and Van Greevenbroek67,Reference Gómez-Banoy, Guseh and Li68)} . Studies have reported elevations of alternative pathway proteins with obesity^{(Reference Pomeroy, Mitchell and Eckert69)} and C3 is positively associated with VAT and SAT^{(Reference Gabrielsson, Johansson and Lönn61)}. Adipocytes also express receptors for C3a and C5a (C3aR and C5aR, respectively) which are produced upon complement activation^{(Reference Phieler, Chung and Chatzigeorgiou70)}. These findings suggest complement activation may increasingly act locally, at the adipose tissue, in obesity. This may promote a pro-inflammatory environment and exert deleterious consequences on adipose tissue biology^{(Reference Phieler, Chung and Chatzigeorgiou70)}.

Beyond the inflammatory influence of complement, its involvement in macronutrient metabolism has been recognised. During the postprandial period, transient elevations in C3 have been reported and are correlated with postprandial TAG concentration^{(Reference Van Oostrom, Alipour and Plokker71,Reference Halkes, Van Dijk and De Jaegere72)} ; there is also evidence of increased complement activation, particularly of the alternative pathway^{(Reference Shim, Begum and Yang60,Reference Peake, Kriketos and Campbell73)} . Fujita et al. found a dose–response relationship between chylomicrons and alternative pathway activation in vitro ^{(Reference Fujita, Fujioka and Murakami74)}, suggesting the consumption of dietary fat may stimulate complement by this pathway. In the alternative pathway, the interaction of factors B and D with C3 generates C3a which is rapidly cleaved to produce C3-acylation-stimulating protein (ASP). ASP promotes TAG synthesis and glucose transport, whilst reducing lipolysis in healthy adipocytes^{(Reference Pattrick, Luckett and Yue75,Reference MacLaren, Cui and Cianflone76)} . Obesity is associated with elevated fasting ASP^{(Reference Maslowska, Phelis and Sniderman77)}, which is positively correlated with postprandial TAG concentration^{(Reference Cianflone, Zakarian and Couillard78)}. These findings suggest that dysregulation of ASP production and signalling (termed ASP resistance) may promote reduced adipose tissue TAG clearance^{(Reference Hertle, Stehouwer and Van Greevenbroek67)}. Furthermore, Xin et al. reported an association between complement components and an adverse lipoprotein profile, with significant associations between C3 and a greater number and particle size of VLDL, greater number and smaller particle size of intermediate-density lipoprotein and LDL, and fewer and smaller HDL particles; the authors suggest that binding of C3 to lipoproteins may negatively influence their metabolism^{(Reference Xin, Hertle and van der Kallen79)}. These findings suggest an emerging role of complement activation in the development of adipose dysfunction, inflammation and the promotion of a pro-atherogenic lipoprotein phenotype, all of which are central defects in the pathophysiology of cardiometabolic disease.

Obesity and body composition

Considering the high rates of cardiometabolic disease in BAC and the strong association between obesity and disease risk, high rates of obesity may also be expected in BAC populations. In a recent UK survey, BAC children, adolescents and adults indeed suffered a high prevalence of obesity⁽⁴⁾. However, earlier work has highlighted pronounced sex differences in obesity within BAC populations, with an alarmingly high prevalence in women, but rates of obesity in men which are similar to that of WE^{(Reference Wang, McPherson and Marsh80,81)} . This pattern is also evident in the USA^{(Reference Wang, McPherson and Marsh80)}. Interestingly, in both sexes, data modelled from the UK Biobank revealed that BAC with a BMI of 26 kg/m² experience an equivalent risk of type 2 diabetes as WE with a BMI of 30 kg/m^{2(Reference Ntuk, Gill and Mackay82)}. This may suggest BAC populations are more sensitive to the effects of excess adiposity. However, weaker associations between BMI and cardiometabolic risk factors have also been reported in BAC^{(Reference Taylor, Coady and Levy83)}, suggesting factors outside of adiposity may be more important in the development of cardiometabolic disease in this population. This hypothesis is supported by a higher probability of BAC being diagnosed with type 2 diabetes in the normal and overweight BMI categories^{(Reference Paul, Owusu Adjah and Samanta7)}.

VAT is associated with cardiometabolic risk independently of BMI^{(Reference Hanley, Wagenknecht and Norris14,Reference Smith, Borel and Nazare15)} . Ethnic differences in body composition are well studied and have revealed significantly lower VAT alongside similar or greater SAT in BAC populations, regardless of sex^{(Reference Carroll, Fulda and Chiapa84–Reference Goedecke, Levitt and Lambert87)}. This gives rise to a more beneficial VAT:SAT ratio, which suggests BAC populations preferentially store excess fatty acids in SAT. In line with these observations, BAC women have been found to exhibit a lower increase in VAT per unit of waist circumference, raising concerns regarding the use of waist circumference to predict VAT in BAC populations^{(Reference Sumner, Micklesfield and Ricks88)}. The consequence of this is that whilst VAT is associated with most cardiometabolic risk factors in WE, in BAC, these associations are weaker or non-existent^{(Reference Rønn, Andersen and Lauritzen89)}. If BAC have a more favourable VAT:SAT ratio, why are there such high rate of cardiometabolic disease in these populations? What is the relevance of VAT in disease pathophysiology in BAC?

Ectopic lipid

According to the theory of ectopic lipid deposition, lower VAT in BAC populations would be expected to drive lower ectopic lipid deposition. To investigate this hypothesis, we conducted a systematic review and meta-analysis to compare IHL, IMCL and IPL deposition in BAC compared to other ethnic populations^{(Reference Reed, Nevitt and Kemp90)}. We found strong evidence for lower IHL in BAC compared to WE, Hispanics and south Asian populations, which was supported by meta-analyses; these differences held regardless of sex, age, BMI and glycaemic status^{(Reference Reed, Nevitt and Kemp90)}. These findings are in line with observations in adolescents^{(Reference Lee and Kuk91,Reference Liska, Dufour and Zern92)} and the observation of markedly lower rates of non-alcoholic fatty liver disease in BAC compared to other ethnic groups^{(Reference Rich, Oji and Mufti93)}. The mechanism(s) driving lower IHL accumulation in BAC are not well understood. In line with the portal theory^{(Reference Lewis, Carpentier and Adeli28)}, lower VAT in BAC appears to translate to lower IHL. However, whilst VAT is significantly associated with IHL in WE, neither VAT nor SAT is associated with IHL in BAC^{(Reference Thamer, Machann and Haap43,Reference Hakim, Bello and Ladwa86)} . This suggests that fatty acid pathways other than adipose tissue lipolysis make a larger contribution to IHL deposition in BAC. The accumulation of IHL and the promotion of hepatic insulin resistance are postulated to be primary defects in the development of cardiometabolic disease^{(Reference Fabbrini, Magkos and Mohammed46)}, therefore lower IHL in BAC populations is paradoxical to their high rates of disease. Interestingly, both BAC and WE women exhibit significant negative associations between IHL and hepatic insulin sensitivity^{(Reference Goedecke, Keswell and Weinreich94,Reference Chung, Courville and Onuzuruike95)} , but no such associations have been reported in BAC men^{(Reference Hakim, Bello and Bonadonna96,Reference Ladwa, Bello and Hakim97)} . These contrasting findings suggest BAC women may be more sensitive to the lipotoxic effects of IHL accumulation, whereas factors other than IHL promote cardiometabolic disease in BAC men. However, this requires further investigation.

Less research has been conducted into ethnic differences of IMCL. The majority of studies report no ethnic differences between BAC and WE (six out of eight studies identified in our systematic review)^{(Reference Reed, Nevitt and Kemp90)}. However, few studies account for ethnic differences in muscle fibre type, of which BAC are characterised by more type 2 and less type 1 fibres^{(Reference Tanner, Barakat and Dohm98)}; a less oxidative phenotype may influence their propensity to IMCL deposition and future studies should account for this variable. In BAC, studies have failed to find an association between IMCL and peripheral insulin sensitivity regardless of sex^{(Reference Bello, Ladwa and Hakim99,Reference Smith, Yao-Borengasser and Starks100)} , which suggests factors other than IMCL are promoting peripheral insulin resistance in this population. Whilst IMCL has been found to be associated with peripheral insulin resistance in WE^{(Reference Anderwald, Bernroider and Krssák53)}, lipid metabolites such as diacylglycerol and ceramides are thought to be more important in the promotion of insulin resistance that total IMCL^{(Reference Samuel and Shulman101)}. Therefore, future studies should focus on ethnic differences in lipid metabolites, and their role in peripheral insulin resistance in BAC.

Whilst few studies have investigated ethnic differences in IPL, those that have done so predominantly report lower IPL in BAC than WE^{(Reference Reed, Nevitt and Kemp90)}. In the twin-cycle hypothesis, it is postulated that elevations in IHL drive IPL deposition via increased VLDL–TAG output^{(Reference Taylor49)}. Therefore, it may be hypothesised that lower IHL is driving lower IPL in BAC^{(Reference Hakim, Bello and Ladwa86)}. However, more studies are needed to clarify ethnic differences in the mechanisms of IPL accumulation and its influence on markers of β-cell function in BAC populations.

Fatty acid metabolism

Ectopic lipid deposition is proposed to be a consequence of deleterious fatty acid metabolism, driven by aberrant adipose tissue function^{(Reference Lewis, Carpentier and Adeli28,Reference Taylor49)} . Therefore, exploring the pathways leading to lower IHL and IPL in BAC populations is of interest. BAC populations have a more beneficial fasting lipid profile, with lower TAG and LDL, alongside higher HDL^{(Reference Zoratti102,Reference Zoratti, Godsland and Chaturvedi103)} . BAC populations are also reported to exhibit lower small-dense LDL compared to WE^{(Reference Hoogeveen, Gaubatz and Sun104)}. Whilst this profile is in line with the lower CHD rates experienced by this population, it remains paradoxical to their high risk of type 2 diabetes and ischaemic stroke^{(4,Reference Tillin, Hughes and Mayet5)} . Interestingly, in the USA, there is evidence for BAC populations losing their cardio-protective fasting lipid profile and levels of HDL do not appear to differ to those in WE^{(Reference Woudberg, Goedecke and Lecour105)}; these findings may represent progressive acculturation to a western nutritional and lifestyle environment.

Despite the established ethnic differences in fasting lipid profiles, little is understood about postprandial fatty acid handling in BAC. In the few studies investigating ethnic differences in postprandial fatty acid dynamics, lower total postprandial TAG has been reported in BAC compared to WE women^{(Reference Punyadeera, Crowther and van der Merwe106–Reference Bower, Deshaies and Pfeifer108)}, and a single study revealed lower total postprandial TAG in young BAC compared to WE men^{(Reference Friday, Srinivasan and Elkasabany109)}. Findings of lower postprandial TAG concentrations in BAC are in line with lower ectopic lipid deposition and a more beneficial fasting lipid profile in this population^{(Reference Reed, Nevitt and Kemp90,Reference Zoratti102,Reference Zoratti, Godsland and Chaturvedi103)} , yet remain paradoxical to the their high rates of cardiometabolic disease^{(4,Reference Tillin, Hughes and Mayet5)} . In studies investigating ethnic differences in the incremental TAG response to feeding (adjusting for baseline TAG which is consistently lower in BAC populations) conflicting findings have been produced. Lower^{(Reference Bower, Deshaies and Pfeifer108)}, as well as equivalent^{(Reference Bower, Deshaies and Pfeifer108,Reference Muniyappa, Sachdev and Sidenko110)} , postprandial TAG increments have been reported in BAC compared to WE women, whereas both lower^{(Reference Friday, Srinivasan and Elkasabany109)} and higher^{(Reference Goff, Whyte and Samuel111)} increments have been reported in BAC men. We have utilised stable isotope techniques to conduct in depth investigations of postprandial fatty acid trafficking between overweight and obese, but otherwise healthy, BAC and WE men. Stable isotope techniques are considered the gold standard for lipid and lipoprotein research, enabling the differentiation between endogenous and meal-derived fatty acids^{(Reference Umpleby112,Reference Borén, Taskinen and Adiels113)} . In response to consecutive moderate-high fat meals, with the first of these meals containing a U–¹³C palmitate stable isotope tracer to label meal-derived fatty acids, we observed a trend for a greater plasma TAG tracer:tracee ratio in BAC compared to WE men^{(Reference Reed, Shojaee-Moradie and Fielding114)}, suggesting that meal-derived fatty acids (transported predominantly in chylomicrons) make a greater relative contribution to total postprandial TAG concentrations in BAC. Whether there are ethnic differences in chylomicron–TAG metabolism requires further research, however lower fasting VLDL–TAG has been reported in BAC men and women^{(Reference Miller, Patterson and Okunade115–Reference Johnson, Slentz and Duscha117)}. Considering the association between IHL and VLDL–TAG in WE^{(Reference Adiels, Taskinen and Packard48)}, these findings are in line with lower IHL typically observed in this population^{(Reference Reed, Nevitt and Kemp90)}. Therefore, the greater tracer:tracee ratio in BAC men may be explained by lower VLDL–TAG rather than elevated chylomicron–TAG, however this has yet to be directly compared in the postprandial period.

Postprandial TAG concentration is determined by both the production and clearance of TAG-rich lipoproteins. VLDL–TAG and chylomicron–TAG compete for LPL affinity, however it appears large TAG-rich chylomicrons are preferentially hydrolysed^{(Reference Bickerton, Roberts and Fielding118)}. If BAC populations exhibit lower VLDL–TAG during the postprandial period, this would imply less competition for LPL affinity and may allow for more efficient chylomicron–TAG clearance. Furthermore, BAC populations have been found to exhibit greater post-heparin LPL activity^{(Reference Bower, Deshaies and Pfeifer108,Reference Friday, Srinivasan and Elkasabany109,Reference Després, Couillard and Gagnon119)} and higher LPL expression in SAT^{(Reference Bower, Deshaies and Pfeifer108)}. Taken together, these findings suggest BAC populations may have a greater capacity for TAG clearance, which is in line with their more beneficial VAT:SAT ratio^{(Reference Carroll, Fulda and Chiapa84–Reference Goedecke, Levitt and Lambert87)}. In our own work, we reported a lower concentration of meal-derived U–¹³C palmitate in TAG at a density of Svedberg floatation rate (Sf) 20–400 (which approximates VLDL–TAG)^{(Reference Reed, Shojaee-Moradie and Fielding120)}. Whilst meal-derived TAG is predominantly transported in chylomicrons, during the late postprandial period, meal-derived TAG may enter the hepatic VLDL–TAG pool via uptake of meal-derived NEFA spillover and chylomicron–remnant-TAG^{(Reference Hodson and Frayn45)}. In BAC, the lower meal-derived U–¹³C palmitate in TAG within Sf 20–400 TAG suggests that meal-derived fatty acids are being cleared by adipose tissue, preventing hepatic uptake and incorporation into VLDL–TAG. However, this may also be influenced by ethnic differences in hepatic partitioning of meal-derived fatty acids, favouring storage and/or oxidation, rather than VLDL–TAG export^{(Reference Hodson and Frayn45)}. Further studies that combine stable isotope and arteriovenous difference techniques are required to elucidate ethnic differences in adipose tissue TAG clearance in BAC populations.

Inflammation and complement

Obesity is associated with a state of chronic low-grade inflammation; however obese BAC populations exhibit a specific inflammatory profile. Adiponectin is consistently reported to be lower in BAC compared to WE^{(Reference Hakim, Bello and Ladwa121,Reference Hyatt, Phadke and Hunter122)} . Described as an insulin sensitiser, adiponectin is significantly associated with insulin sensitivity in WE; however an association between adiponectin and insulin sensitivity is not found in BAC men^{(Reference Hakim, Bello and Ladwa123)}, which questions its metabolic function in this population. BAC also exhibit lower TNF-α^{(Reference Hakim, Bello and Ladwa121,Reference Hyatt, Phadke and Hunter122,Reference Beasley, Koster and Newman124)} , alongside higher or similar IL-10^{(Reference Hakim, Bello and Ladwa121,Reference Beasley, Koster and Newman124)} , suggesting a relatively more anti-inflammatory profile compared to WE. Similarly, associations between pro-inflammatory cytokines (TNF-α and IL-6) and insulin sensitivity appear to be weaker in BAC compared to WE^{(Reference Hakim, Bello and Ladwa123,Reference Beasley, Koster and Newman124)} . Ethnic differences in the associations between inflammatory markers and adiposity have also been reported. Adiponectin is significantly associated with SAT in WE but not BAC, and IL-6 is associated with VAT and SAT in WE but not BAC^{(Reference Hakim, Bello and Ladwa121)}. Taken together, this implies a specific inflammatory profile in BAC, with lower adiponectin and TNF-α and greater IL-6. There also appear to be differences in the associations between adipose tissue depots and insulin sensitivity, suggesting differences in obesity-induced chronic low-grade inflammation and its role in cardiometabolic disease.

Very little is known about the complement system in BAC populations, despite its emerging role in cardiometabolic disease^{(Reference Gabrielsson, Johansson and Lönn61–Reference Wlazlo, Van Greevenbroek and Ferreira63)}. The potential for ethnic differences in complement were first considered in response to COVID-19, in which BAC populations suffered one of the highest mortality rates of all ethnic groups^{(Reference Aldridge, Lewer and Katikireddi125)}. Following the identification of complement dysregulation in severe COVID-19^{(Reference Chauhan, Wiffen and Brown126)}, it was hypothesised that BAC may exhibit greater complement dysregulation, predisposing this population to severe COVID-19, but also to high rates of cardiometabolic disease. In the first ethnic comparison of fasting circulating complement markers, Goff et al.^{(Reference Goff, Davies and Zelek127)} investigated ethnic differences between BAC and WE men. Age-adjusted C3, a central complement component, as well as C4 of the classical pathway were higher in BAC^{(Reference Goff, Davies and Zelek127)}. C3 and C4 are the most abundant complement proteins, and elevated concentrations may suggest a greater capacity for complement activation. BAC also exhibited significantly greater iC3b, indicating greater upstream activation of C3. Additionally, there were significant ethnic differences in markers of the alternative pathway, with BAC exhibiting lower factor D and higher properdin compared to WE^{(Reference Goff, Davies and Zelek127)}. Factor D is enzyme primarily secreted by adipose tissue that activates the alternative pathway by cleaving factor B; properdin is also expressed in adipose tissue and stabilises the C3 convertase complex^{(Reference Hertle, Stehouwer and Van Greevenbroek67,Reference Gómez-Banoy, Guseh and Li68)} . Considering BAC have a lower body fat percentage for any given BMI^{(Reference Deurenberg, Yap and Van Staveren128)}, ethnic differences in factor D may be attributed to lower fat mass in BAC. However, greater concentrations of properdin in this population may suggest a greater propensity for C3 cleavage (by C3 convertase), driving greater downstream complement activation. These findings suggest greater complement dysregulation in BAC, which are in line with reports of a higher prevalence of genetic variants associated with regulators promoting complement activation in BAC^{(Reference Zhao, Wu and Khosravi129)}.

Considering the positive association between complement markers and insulin resistance/type 2 diabetes^{(Reference Wlazlo, Van Greevenbroek and Ferreira63)}, it may be that complement dysregulation is an important contributor to the high rates of type 2 diabetes observed in BAC populations. In support of this, we revealed an association between HbA1c and C3 independent of adiposity in BAC, but not WE^{(Reference Reed, Hakim and Lockhart130)}. Understanding ethnic differences in the role of complement in adipose tissue function, lipid metabolism and inflammation in BAC populations is an important avenue of research, particularly in the postprandial period. Complement dysregulation is also associated with an adverse lipoprotein profile and atherosclerotic CVD^{(Reference Hertle, van Greevenbroek and Arts62,Reference Xin, Hertle and van der Kallen79)} . Interestingly, despite exhibiting greater complement dysregulation, BAC participants are recognised to have a more beneficial fasting lipid profile^{(Reference Zoratti102–Reference Hoogeveen, Gaubatz and Sun104)}. Therefore, there may be ethnic differences in the role of complement in CVD in BAC populations. Exploration of the mechanisms driving complement dysregulation in BAC and their role in cardiometabolic disease requires further investigation.