Table 1: Summary of characteristics of the included studies: methods, population of interest, glucose tolerance status, findings, and mNOS.

StudyMethodsStudy designPopulation of interestComparison populationFindingsmNOS
EthnicityMean ageGender %M/FGlucose tolerance

Rubenstein et al. [24]OGTTCross-sectionalNative black South African, no FHx of T2D7100/0NGTWhite South African, no FHx of T2D8Similar fasting insulin levels, but lower serum insulin in response to oral glucose and lower renal insulin clearance in BA compared to WE5
Shires et al. [25]“Maximal beta cell stimulation”—75 g oral glucose followed by IV tolbutamide and glucagon stimulationCross-sectionalObese native black South African100/100NondiabeticObese white South African8Serum insulin and C-peptide levels significantly lower in BA compared to WE 30 mins after oral glucose and 30 mins after tolbutamide and glucagon6
Cruickshank et al. [31]OGTTCross-sectionalUK Afro-Caribbean106 (F), (M)50/50NGT, IGT, and T2DWhite European101No statistically significant difference between C-peptide and insulin concentrations post oral glucose; however, difference in profiles led investigators to suggest impaired hepatic processing of insulin in BA.6
Osei et al. [48]OGTTCase controlAfrican American (first-degree relatives of T2D and controls)24 relatives and 8 controls
(relatives) and (controls)0/100NGTWhite American (first-degree relatives of T2D and controls)22In both first-degree relatives and controls, iAUC insulin and C-peptide were significantly higher in BA compared to WE.9
UK Prospective Diabetes Study Group [32]Fasting glucose and insulin to derive HOMA%BProspective cohortUK Afro-Caribbean387 (M)
(F)
57/43T2DUK Caucasian4177Lower HOMA%B in BA compared to WE, adjusted for age and BMI. BA had higher fasting plasma glucose and higher HbA1c vs. WE.9
Osei et al. [63]Tolbutamide-modified IVGTTCross-sectionalAfrican American320/100NGTWhite American30Higher insulin levels in BA but similar C-peptide levels compared to WE9
Haffner et al. [39]Insulin-modified IVGTTProspective cohortAfrican American187 NGT, 101 IGT
43/57NGT and IGTNon-Hispanic white229Higher AIR in BA compared to WE when adjusted for age, sex, BMI, and WHR. Greater insulin resistance in BA by Si9
Haffner et al. [38]OGTT
Insulin-modified IVGTT
Prospective cohortAfrican American (overweight and obese)15345/55T2DNon-Hispanic white (overweight and obese)105OGTT: no difference in fasting and 2 h insulin. AIR higher in BA. No ethnic difference in Si. Adjusted for age, sex, obesity and WHR, fasting glucose, and therapy9
Osei et al. [34]OGTTCross-sectionalNative Ghanaian, US immigrant Ghanaian, and African American50, 31, and 66
, , and 39/61NGTWhite American39Higher peak and iAUC insulin in BA compared to WE, no difference in C-peptide. Basal and postprandial HIE (by C-peptide to insulin molar ratio) lower in BA9
Chiu et al. [64]Hyperglycaemic clampCross-sectionalAfrican American92656/44NGTWhite American34Higher second-phase insulin response in BA, no difference in the 1st phase. Lower insulin sensitivity index in BA7
Melby et al. [49]Insulin-modified IVGTTCross-sectionalAfrican American90/100NGTWhite American8Age-matched ethnic groups. Higher AIR in BA and lower insulin sensitivity by Si in BA. Similar age, BMI, and percentage body fat between ethnicities9
Chiu et al. [65]Hyperglycaemic clampCross-sectionalAfrican American112545/55NGTWhite American46Higher 1st and 2nd phase insulin response in BA, lower insulin sensitivity index6
Chen et al. [50]Glucose challenge test (insulin and C-peptide 1-hour post 50 g glucose load)Prospective cohortAfrican American pregnant women3430/100NGTWhite American pregnant women118Insulin and insulin to glucose ratio higher, but C-peptide and C-peptide to insulin ratio lower, in BA vs. WE. Similar BMI in both ethnic groups8
Jensen et al. [40]OGTTCohortAfrican American5538/62NGT, IFG/IGT, and T2DWhite American217Insulinogenic index adjusted for insulin sensitivity ((ΔI30/ΔG30)/HOMA-IR) higher in NGT BA vs. WE. No difference in IFG/IGT and T2D7
Punyadeera et al. [26]MMTTCross-sectionalNative black South African (lean and obese)8 (lean) and 9 (obese)
(lean) and (obese)0/100NGTWhite South African (lean and obese)17Higher insulin levels at 30 mins post meal in obese BA compared to obese WE. No difference in total insulin area under the curve between ethnic groups8
Velasquez-Mieyer et al. [42]3 h OGTTCross-sectionalAfrican American (severely obese)16Not available by ethnicityNGT and IGTWhite American (severely obese)26Insulin iAUC and CIR at 30 mins higher in BA vs. WE, no significant difference in iAUC C-peptide. Similar WBISI and fat mass by DXA between ethnic groups8
Stefan et al. [59]OGTT, MMTT, and IVGTTCross-sectionalAfrican American3066/33NGTWhite American30No ethnic difference in AUC insulin post OGTT or MMTT. Higher AIR after IVGTT in BA, when adjusted for insulin sensitivity by M-Low (from 2-step hyperinsulinaemic-euglycaemic clamp). BA and WE matched for age, sex, BMI, WHR, and percentage body fat.9
Torrens et al. [41]Fasting glucose and insulin to derive HOMA%BProspective cohortAfrican American7460/100Nondiabetic; included subjects with IFGNon-Hispanic white1359HOMA%B higher in BA after controlling for alcohol consumption, waist circumference, triglycerides, and prevalence of IFG8
Albu et al. [43]Insulin- or tolbutamide-modified IVGTTCross-sectionalAfrican American320/100NGT and IGTWhite American28Higher AIR in BA, which persisted when adjusted for insulin sensitivity index and measures of adiposity and skeletal muscle7
Boule et al. [61]IVGTTPre/postinterventionalAfrican American17332.9 (30.8-35.0) (F)
33.1 (30.2-36.0) (M)
37/63NondiabeticWhite American423AIRg and DI higher and Si lower in BA vs WE8
Reimann et al. [28]OGTTCross-sectionalBlack South African (with and without abdominal obesity)8627 (23, 30) (without abdominal obesity)
32 (28, 35) (with abdominal obesity)
0/100NondiabeticWhite South African (with and without abdominal obesity)90No ethnic difference in age, BMI, or insulin resistance by HOMA-IR. 2 h C-peptide post glucose significantly higher in BA vs. WE in the group with abdominal obesity. No ethnic difference in post glucose C-peptide in the lean group7
Elbein et al. [44]OGTT
Insulin- or tolbutamide-modified IVGTT
Case controlAfrican American15934/66NGT and IGTEuropean American344OGTT: area under curve insulin and insulinogenic index higher in BA. AIRg and DI higher in BA, no significant difference in Si9
Herman et al. [35]OGTTRCT (data obtained before randomisation)African American75226/74IGTWhite American2117Higher CIR in BA in the context of greater insulin resistance in BA by HOMA IR and higher BMI7
Rasouli et al. [60]OGTT
IVGTT
Arginine stimulation test in subgroup
Cross-sectionalAfrican American (obese, with and without family hx of T2D)12138 (36-39)41/59NGTWhite American (obese, with and without family hx of T2D)212No difference in 2 h OGTT insulin or insulin area under the curve, but higher AIRg in BA. No difference in AIRmax, lower DI max. BA had lower Si but higher disposition index.
No difference in age, BMI, or WHR between ethnic groups
9
Goedecke et al. [27]OGTT
Insulin-modified IVGTT
Cross-sectionalNative black South African (lean and obese)29 (lean) and (obese)0/100NGTWhite South African (lean and obese)28Higher insulinogenic index 30 mins in BA, but no difference when adjusted for insulin sensitivity. Higher AIR and lower Si in BA, adjusted for visceral and subcut adipose volume. No ethnic difference in DI. No ethnic difference in BMI, WHR, or body fat % by DXA9
Willig et al. [51]Insulin- or tolbutamide-modified IVGTTCross-sectionalAfrican American870/100NondiabeticEuropean American68AIRg higher and Si lower in BA. No ethnic difference in age and BMI8
Chandler-Laney et al. [52]Insulin-modified IVGTTCross-sectionalAfrican American43 (premenopausal), (postmenopausal)0/100NGTEuropean American63Higher X0 (acute C-peptide secretion), Phi1 and Phi(tot) in BA after adjusting for age. Lower Si in BA, after adjusting for body fat % by DXA9
Goree et al. [53]Insulin-modified IVGTTCross-sectionalAfrican American42Premenopausal
Postmenopausal
0/100NGTEuropean American64AIRg and DI higher in BA. AIR remained higher after adjusting for Si.9
Chow et al. [54]Insulin-modified IVGTTCross-sectionalAfrican American170/100NondiabeticWhite American17AIRg and DI higher in BA, Si lower. Matched for age and BMI. Lower WHC and higher body fat % in BA7
Ladson et al. [55]OGTTCase controlAfrican American with PCOS360/100NondiabeticWhite American with PCOS63Higher fasting insulin and insulinogenic index (30 mins) in BA. Similar BMI and WHR in both ethnic groups6
Szczepaniak et al. [45]Insulin-modified IVGTTCross-sectionalAfrican American2035/65NGT and IGTNon-Hispanic white30Higher AIRg and DI in BA, lower Si. Ethnic groups matched for age, sex, BMI, and BP. Adjusted for HbA1c8
Goff et al. [33]IVGTTCross-sectionalUK Afro-Caribbean35 (F)
(M)
29/71NondiabeticUK white417When adjusted for age and BMI: higher AIRg and lower Si in BA. No ethnic difference in DI7
Ebenibo et al. [62]IVGTTProspective cohortAfrican American142 (), (HbA1c 5.7-6.4%)25/75NGTWhite American138Within each HbA1c group, BA had higher AIRg and DI. Similar insulin sensitivity between ethnic groups by hyperinsulinaemic-euglycaemic clamp9
Ferguson et al. [58]HOMA%B
Insulin-modified IVGTT
Pre/postinterventionalAfrican American4226 (median), 9 (IQR) (F)
27 (median), 18 (IQR) (M)
45/55NondiabeticEuropean American106Higher HOMA%B in BA
Lower Si and higher AIR and DI in BA, adjusted for age, sex, and BMI
8
Healy et al. [36]HOMA%B
OGTT
Insulin-modified IVGTT
Cross-sectionalAfrican American (obese)847/93Prediabetic (IFG and IGT)White American (obese)61HOMA%B lower in BA
OGTT fasting and post glucose C-peptide levels lower in BA, no difference in insulinogenic index
Higher AIRg in BA (not significant), similar Si, and significantly higher DI in BA. BA had higher BMI and higher body fat % by DXA.
9
Goff et al. [29]MMTT plus high-fructoseor high-glucose feedingCross-sectionalUK black African9100/0NondiabeticUK white417No significant ethnic difference in insulin or C-peptide iAUC post feeding7
Owei et al. [46]IVGTTProspective cohortAfrican American (with parental T2D)184Not reported by ethnicityNGT and IGTEuropean American (with parental T2D)151Higher AIR in BA. Lower insulin sensitivity by hyperinsulinaemic-euglycaemic clamp in blacks. BA younger and higher BMI8
Shah et al. [47]Hyperglycaemic clamp
Arginine-stimulated insulin response
Cross-sectionalAfrican American24Not reported by ethnicityNot reported by ethnicityNGT, IGT, and T2DWhite American74Higher acute insulin response and DI in BA, late-phase insulin tended to be higher in BA. No ethnic difference in insulin sensitivity index8
Osei et al. [37]Fasting parameters to determine HOMA%B
OGTT
Insulin-modified IVGTT
Cross-sectionalOverweight/obese African American670/100Prediabetic (IFG and IGT)Overweight/obese white American28HOMA%B: no ethnic difference
OGTT: fasting C-peptide and peak C-peptide lower in BA, fasting and mean insulin tended to be higher
IVGTT: AIR higher in BA, not significant. DI significantly higher in BA. BA had higher BMI and higher percentage body fat by DEXA. Insulin sensitivity by Si same between ethnic groups
9
Piccinini et al. [56]Insulin-modified IVGTTCross-sectionalAfrican American180/100NGTEuropean American29Insulin secretion rate (ISR) as modelled by C-peptide higher in BA vs. WE8
Mohandas et al. [30]MMTT
Hyperglycaemic clamp
Cross-sectionalUK black African19100/0T2D (recently diagnosed)UK white15MMTT: fasting and AUC C-peptide lower in BA, no difference in insulin AUC
HC: second-phase C-peptide lower in BA, no difference in insulin iAUC. Groups matched for age, BMI, HbA1c, and duration of diabetes
9

AIR: acute insulin response; BA: black African; BMI: body mass index; CIR: corrected insulin response; DI: disposition index; DXA: dual-energy X-ray absorptiometry; FHx: family history; HIE: hepatic insulin extraction; HOMA%B: homeostatic model assessment of beta cell function; HOMA-IR: homeostatic model assessment of insulin resistance; iAUC: incremental area under the curve; IGT: impaired glucose tolerance; IVGTT: intravenous glucose tolerance test; mNOS: modified Newcastle-Ottawa scale; MMTT: mixed meal tolerance test; NGT: normal glucose tolerance; OGTT: oral glucose tolerance test (refers to 2-hour post 75 g oral glucose); PCOS: polycystic ovarian syndrome; RCT: randomised controlled trial; Si: insulin sensitivity index; T2D: type 2 diabetes; WBISI: whole-body insulin sensitivity index; WE: white European; WHR: waist-hip ratio.