Identification of a subgroup of black South Africans with type 1 diabetes who are older at diagnosis but have lower levels of glutamic acid decarboxylase and islet antigen 2 autoantibodies.

AIMS: To compare the age at diagnosis and prevalence of islet autoantibody [glutamic acid decarboxylase (65 kDa) 65 and islet antigen 2] positivity in black and white participants with type 1 diabetes in South Africa, and to analyse the relationship between age at diagnosis and the presence of autoantibodies. METHODS: Participants were recruited from diabetes outpatient departments and autoantibodies to glutamic acid decarboxylase (65 kDa) and islet antigen 2 were measured by enzyme-linked immunosorbent assay. RESULTS: We recruited 472 (353 black and 119 white) participants with type 1 diabetes. Age at diagnosis of diabetes was later in black (19.7 ± 10.5) than in white participants (12.7 ± 10.8 years; P < 0.001) with a median (interquartile range) disease duration of 5.0 (2.0-10.0) and 8.5 (4.0-20.0) years (P < 0.001), respectively. An older age at diagnosis (≥ 21 years) was more frequent in black (152 of 340, 45%) than in white participants (24 of 116, 21%; P < 0.001). The prevalence of islet antigen 2 autoantibodies was 19% (66/352) in black and 41% in white participants (48/118; P < 0.001). There was no significant difference in glutamic acid decarboxylase (65 kDa) autoantibody positivity between black (212/353, 60%) and white participants (77/117, 66%; P = 0.269). In black, but not white, participants the prevalence of both glutamic acid decarboxylase (65 kDa) and islet antigen 2 autoantibody positivity was significantly lower in participants diagnosed at age ≥ 21 years (P < 0.001 for both comparisons). CONCLUSIONS: The older age at diagnosis, lower prevalence of islet antigen 2 autoantibodies and a distinct subgroup of participants with type 1 diabetes with age at diagnosis of > 20 years in the black compared to white population suggest a difference in the immunological aetiology of type 1 diabetes in these two population groups.

Prevalence of low serum testosterone levels among men with type 2 diabetes mellitus attending two outpatient diabetes clinics in KwaZulu-Natal Province, South Africa.

BACKGROUND: The reported prevalence of low testosterone among men with type 2 diabetes mellitus (T2DM) is high. However, there is a dearth of information on the prevalence of androgen deficiency symptoms and low serum testosterone levels in men with T2DM from sub-Saharan Africa. Scanty data are available from Nigeria, Ghana and South Africa (SA). OBJECTIVES: To determine the prevalence of low serum testosterone and associated risk factors and the prevalence of androgen deficiency symptoms in men with T2DM. METHODS: In a cross-sectional observational study, androgen deficiency symptoms in men with T2DM attending two outpatient diabetes clinics in Durban, KwaZulu-Natal Province, SA, were assessed using the Ageing Males' Symptoms Scale (AMS) questionnaire and direct enquiry. Serum total testosterone (TT), sex hormone-binding globulin (SHBG), luteinising hormone (LH), fructosamine, serum lipids and glycated haemoglobin (HbA1c) were measured and free testosterone (FT) was calculated. TT, SHBG and FT levels were measured in control subjects with no history of diabetes. RESULTS: There were 148 men with T2DM in the study group and 50 control subjects in the control group. In the study group, the majority were black Africans (58.8%); Indians (39.2%) and whites (2.0%) constituted the remainder. The mean (standard deviation (SD)) age was 57.5 (11.2) years, the mean duration of diabetes 11.4 (8.9) years and the mean HbA1c 8.6% (1.9%). Of the study group, 85.8% had metabolic syndrome. Mean TT, SHBG and FT and median LH (interquartile range) in the study group were within normal ranges. However, mean (SD) serum TT and FT were lower in the study group than in the control subjects (14.5 (5.8) v. 18.8 (7.2) nmol/L; p<0.001 and 265.9 (90.4) v. 351.7 (127.3) pmol/L; p<0.001, respectively). The prevalence of low serum total testosterone (LSTT) and low serum free testosterone (LSFT) in the study group was 35.8% and 16.2%, respectively. The prevalence of androgen deficiency symptoms using the AMS questionnaire was 74.5% and correlated poorly with LSTT or LSFT. In multivariate analysis, LSFT was significantly associated with age (odds ratio (OR) 1.05, 95% confidence interval (CI) 1.02 - 1.218; p=0.043) and waist circumference (WC) (OR 1.033, 95% CI 0.999 - 1.068; p=0.059). LSTT was associated with body mass index (BMI) only (OR 1.138, 95% CI 1.063 - 1.218; p<0.0001). TT correlated inversely with BMI, WC and the number of metabolic syndrome criteria. FT correlated inversely with BMI, WC and WHR. CONCLUSIONS: There was a high prevalence of LSTT, LSFT and androgen deficiency symptoms in this study. Serum TT and FT were lower in men with T2DM than in control subjects. Risk factors associated with LSFT or LSTT included higher BMI and WC and older age. The AMS score was a poor predictor of low testosterone. More research is required locally before any screening policy can be recommended.

Cardiovascular disease management in people with diabetes outside North America and Western Europe in 2006 and 2015.

AIM: Optimal treatment of cardiovascular disease is essential to decrease mortality among people with diabetes, but information is limited on how actual treatment relates to guidelines. We analysed changes in therapeutic approaches to anti-hypertensive and lipid-lowering medications in people with Type 2 diabetes from 2006 and 2015. METHODS: Summary data from clinical services in seven countries outside North America and Western Europe were collected for 39 684 people. Each site summarized individual-level data from outpatient medical records for 2006 and 2015. Data included: demographic information, blood pressure (BP), total cholesterol levels and percentage of people taking statins, anti-hypertensive medication (angiotensin-converting enzyme inhibitors, calcium channel blockers, angiotensin II receptor blockers, thiazide diuretics) and antiplatelet drugs. RESULTS: From 2006 to 2015, mean cholesterol levels decreased in six of eight sites (range: -0.5 to -0.2), whereas the proportion with BP levels > 140/90 mmHg increased in seven of eight sites. Decreases in cholesterol paralleled increases in statin use (range: 3.1 to 47.0 percentage points). Overall, utilization of anti-hypertensive medication did not change. However, there was an increase in the use of angiotensin II receptor blockers and a decrease in angiotensin-converting enzyme inhibitors. The percentage of individuals receiving calcium channel blockers and aspirin remained unchanged. CONCLUSIONS: Our findings indicate that control of cholesterol levels improved and coincided with increased use of statins. The percentage of people with BP > 140/90 mmHg was higher in 2015 than in 2006. Hypertension treatment shifted from using angiotensin-converting enzyme inhibitors to angiotensin II receptor blockers. Despite the potentially greater tolerability of angiotensin II receptor blockers, there was no associated improvement in BP levels.

Deriving an optimal threshold of waist circumference for detecting cardiometabolic risk in sub-Saharan Africa.

BACKGROUND: Waist circumference (WC) thresholds derived from western populations continue to be used in sub-Saharan Africa (SSA) despite increasing evidence of ethnic variation in the association between adiposity and cardiometabolic disease and availability of data from African populations. We aimed to derive a SSA-specific optimal WC cut-point for identifying individuals at increased cardiometabolic risk. METHODS: We used individual level cross-sectional data on 24 181 participants aged ⩾15 years from 17 studies conducted between 1990 and 2014 in eight countries in SSA. Receiver operating characteristic curves were used to derive optimal WC cut-points for detecting the presence of at least two components of metabolic syndrome (MS), excluding WC. RESULTS: The optimal WC cut-point was 81.2 cm (95% CI 78.5-83.8 cm) and 81.0 cm (95% CI 79.2-82.8 cm) for men and women, respectively, with comparable accuracy in men and women. Sensitivity was higher in women (64%, 95% CI 63-65) than in men (53%, 95% CI 51-55), and increased with the prevalence of obesity. Having WC above the derived cut-point was associated with a twofold probability of having at least two components of MS (age-adjusted odds ratio 2.6, 95% CI 2.4-2.9, for men and 2.2, 95% CI 2.0-2.3, for women). CONCLUSION: The optimal WC cut-point for identifying men at increased cardiometabolic risk is lower (⩾81.2 cm) than current guidelines (⩾94.0 cm) recommend, and similar to that in women in SSA. Prospective studies are needed to confirm these cut-points based on cardiometabolic outcomes.International Journal of Obesity advance online publication, 31 October 2017; doi:10.1038/ijo.2017.240.

Sustaining diabetes prevention and care interventions: A multiple case study of translational research projects.

BACKGROUND: This study identifies the barriers and enablers for sustainability of interventions in primary and secondary prevention of diabetes. In the context of translational research, sustainability is defined as the continued use of program components and activities for the continued achievement of desirable program and population outcomes. METHODS: In this study, eleven translational research projects, supported by the BRIDGES program of the International Diabetes Federation, were investigated. By theoretically-informed semi-structured interviews and analyses of project reports, qualitative data was collected on the sustainability outcomes and the barriers and enablers. RESULTS: The sustainability outcomes can be grouped in three main areas: (1) sustainability at the intervention site(s); (2) diffusion to the wider community; and (3) replication of the intervention at other site(s). Each of the outcomes has their own set of enablers and barriers, and thus requires consideration for a different sustainability strategy. CONCLUSIONS: This study is the first international study that relates the sustainability outcomes of translational research project to specific barriers and enablers, and develops an evidence-based framework which provides practical advice on how to ensure the sustainability of health interventions.

The need for an integrated approach for chronic disease research and care in Africa.

With the changing distribution of infectious diseases, and an increase in the burden of non-communicable diseases, low- and middle-income countries, including those in Africa, will need to expand their health care capacities to effectively respond to these epidemiological transitions. The interrelated risk factors for chronic infectious and non-communicable diseases and the need for long-term disease management, argue for combined strategies to understand their underlying causes and to design strategies for effective prevention and long-term care. Through multidisciplinary research and implementation partnerships, we advocate an integrated approach for research and healthcare for chronic diseases in Africa.

Study profile: the Durban Diabetes Study (DDS): a platform for chronic disease research.

The Durban Diabetes Study (DDS) is a population-based cross-sectional survey of an urban black population in the eThekwini Municipality (city of Durban) in South Africa. The survey combines health, lifestyle and socioeconomic questionnaire data with standardised biophysical measurements, biomarkers for non-communicable and infectious diseases, and genetic data. Data collection for the study is currently underway and the target sample size is 10 000 participants. The DDS has an established infrastructure for survey fieldwork, data collection and management, sample processing and storage, managed data sharing and consent for re-approaching participants, which can be utilised for further research studies. As such, the DDS represents a rich platform for investigating the distribution, interrelation and aetiology of chronic diseases and their risk factors, which is critical for developing health care policies for disease management and prevention. For data access enquiries please contact the African Partnership for Chronic Disease Research (APCDR) at data@apcdr.org or the corresponding author.

H3Africa multi-centre study of the prevalence and environmental and genetic determinants of type 2 diabetes in sub-Saharan Africa: study protocol.

The burden and aetiology of type 2 diabetes (T2D) and its microvascular complications may be influenced by varying behavioural and lifestyle environments as well as by genetic susceptibility. These aspects of the epidemiology of T2D have not been reliably clarified in sub-Saharan Africa (SSA), highlighting the need for context-specific epidemiological studies with the statistical resolution to inform potential preventative and therapeutic strategies. Therefore, as part of the Human Heredity and Health in Africa (H3Africa) initiative, we designed a multi-site study comprising case collections and population-based surveys at 11 sites in eight countries across SSA. The goal is to recruit up to 6000 T2D participants and 6000 control participants. We will collect questionnaire data, biophysical measurements and biological samples for chronic disease traits, risk factors and genetic data on all study participants. Through integrating epidemiological and genomic techniques, the study provides a framework for assessing the burden, spectrum and environmental and genetic risk factors for T2D and its complications across SSA. With established mechanisms for fieldwork, data and sample collection and management, data-sharing and consent for re-approaching participants, the study will be a resource for future research studies, including longitudinal studies, prospective case ascertainment of incident disease and interventional studies.

Retinopathy in subjects with type 2 diabetes at a tertiary diabetes clinic in Durban, South Africa: Clinical, biochemical and genetic factors.

AIM: To determine the prevalence of clinical and laboratory variables and genetic polymorphisms in association with diabetic retinopathy (DR) in subjects with type 2 diabetes attending a tertiary referral diabetes clinic in Durban, South Africa. METHODS: Cross-sectional study on 292 Indian and African patients with type 2 diabetes (71.5% women). The presence of DR was determined by direct ophthalmoscopy. Clinical and laboratory data were collected and polymorphisms in the NOS3 (rs61722009, rs2070744, rs1799983) and VEGF (rs35569394, rs2010963) genes were determined. RESULTS: DR was present in 113 (39%) subjects. Those with DR were older (60.6 ± 9.6 vs. 55.4 ± 12.9 years, p = 0.005), had longer duration diabetes (18.5 ± 8.8 vs. 11.9 ± 9.2 years, p < 0.0001), higher HbA1c (9.2 ± 1.8 vs. 8.8 ± 1.7%, p = 0.049), serum creatinine (106.3 ± 90.2 vs. 75.2 ± 33.4 µmol/l), triglycerides (2.1 ± 1.2 vs. 1.9 ± 1.6 mmol/l, p = 0.042), proteinuria (72% vs. 28%, p = 0.001), and used more insulin (78% vs. 39% p = 0.0001), anti-hypertensive (95% vs. 80%, p = 0.0003) and lipid-lowering therapy (70% vs. 56%, p = 0.023). There was no association between DR and any of the NOS3 or VEGF gene polymorphisms studied, although there were ethnic differences. After adjustment, diabetes duration (OR 1.05, 95% CI 1.01-1.08), presence of proteinuria (OR 4.15, 95% CI 1.70-10.11) and use of insulin therapy (OR 3.38, 95% CI 1.60-7.12) were associated with DR. CONCLUSION: Hyperglycemia, duration of diabetes and proteinuria are associated with DR in Indian and African patients in South Africa, whereas NOS3 and VEGF gene polymorphisms were not associated with DR.

Variants in PPARG, KCNJ11, TCF7L2, FTO, and HHEX genes in South African subjects of Zulu descent with type 2 diabetes

Polymorphisms in a number of genes have consistently been associated with type 2 diabetes in various Caucasian populations. Little, however, is known of the association between these genetic risk markers and type 2 diabetes in sub-Saharan African subjects. The aim of the current study was to determine the association between common variants in the PPARG, KCNJ11, TCF7L2, FTO and HHEX genes in African (black) subjects of Zulu descent in KwaZulu-Natal, South Africa. The association between type 2 diabetes and rs1801282 (PPARG), rs5215 (KCNJ11), rs12255372 (TCF7L2), rs7903146 (TCF7L2) rs9939609 (FTO) and rs1111875 (HHEX) was determined in 178 South African Zulu subjects and 200 healthy ethnically matched control subjects. rs1801282 (PPARG) and rs5215 (KCNJ11) were not found to be present in either the subjects with type 2 diabetes or the control subjects. No association between rs12255372 (TCF7L2), rs9939609 (FTO) and type 2 diabetes was found. Heterozygosity at rs7903146 (TCF7L2) was associated with type 2 diabetes (odds ratio 1.84, 95% confidence interval: 1.19­2.83, p=0.0035). Decreased frequency of homozygosity for the common allele at rs7903146 (TCF7L2) was observed in subjects with type 2 diabetes (odds ratio 0.54, 95% confidence interval: 0.34­0.84; p=0.0043). There was an increased frequency of C allele homozygosity in subjects with type 2 diabetes at rs1111875 (HHEX), of borderline significance (odds ratio 1.54, 95% confidence interval 0.97­2.44, p=0.052). Subjects with type 2 diabetes harbouring one or more of the risk alleles did not differ from those without genetic variation at the loci studied, with respect to age at diagnosis, blood pressure, body mass index or serum lipid levels. We conclude that risk polymorphisms identified in Caucasian populations are not associated with type 2 diabetes in this group of South African subjects of Zulu descent, with the exception of rs7903146 (TCF7L2). The genetic risk for type 2 diabetes in sub-Saharan African subjects may reside in other, as yet unidentified, genes

Toll-like receptor 3 gene polymorphisms in South African Blacks with type 1 diabetes.

Type 1 diabetes is the consequence of exposure of genetically susceptible individuals to specific environmental precipitants. The innate immune system provides the initial response to exogenous antigen and links with the adaptive immune system. The aim of this study was to assess the role of polymorphisms occurring in the cytoplasmic region of toll-like receptor (TLR) 3 gene and immediate 5' sequence, in subjects of Zulu descent with type 1 diabetes in KwaZulu-Natal, South Africa. Seventy-nine subjects with type 1 diabetes and 74 healthy normal glucose tolerant gender-matched control subjects were studied. Parts of exon 4 and exon 3/intron 3 of the TLR3 gene were studied by polymerase chain reaction, direct sequencing and restriction enzyme digestion with Bts 1. Of the nine polymorphisms studied, a significant association with type 1 diabetes was found for the major allele in the 2593 C/T polymorphism and for the minor alleles in the 2642 C/A and 2690 A/G polymorphisms, which were found to be in complete linkage disequilibrium. Correction of the P-values for the number of alleles studied, however, rendered the results no longer significant. These results suggest that polymorphisms in the TLR3 gene, which is part of the innate immune system, may be associated with type 1 diabetes in this population.

Distribution of HLA class II (DRB1/DQB1) alleles and haplotypes among Bahraini and Lebanese Arabs.

The genetic relationship between Bahraini and Lebanese Arabs in terms of HLA class II (DRB1 and DQB1) gene and haplotype frequencies was investigated in a group of 90 Lebanese and 52 Bahraini Arabs. Subjects of both sexes were unrelated and HLA-DRB1 and DQB1 genes were genotyped using the polymerase chain reaction-sequence specific primer (PCR-SSP) technique. Analysis of the HLA-DRB1 alleles showed that the DRB1*040101 and DRB1*110101 alleles were more common among Lebanese, whereas DRB1*030101, DRB1*130701/1327, and DRB1*160101 alleles were more common among Bahrainis. Similarly, of the 7 HLA-DQB1 alleles analyzed, the presence of DQB1*0201 was higher among Bahrainis, whereas DQB1*030101 was higher among Lebanese. The DRB1*160101-DQB1*050101 (23.08%) and DRB1*030101-DQB1*0201 (21.15%) haplotypes were more frequent among Bahrainis, while the DRB1*110101-DQB1*030101 (56.67%), DRB1*040101-DQB1*0302 (28.89%) and DRB1*040101/DQB1*030101 (25.56%) haplotypes were more frequent in Lebanese subjects. Our results underline significant differences between these two populations in HLA class II distribution, and provide basic information for further studies of MHC heterogeneity among Arab-speaking countries, and as a reference for further anthropologic studies.

High incidence of Type 2 diabetes mellitus in South African Indians: a 10-year follow-up study.

AIMS: Previous cross-sectional studies have established that South African Indians have a high prevalence of Type 2 diabetes mellitus. A prospective community study was undertaken to determine the incidence of Type 2 diabetes and the risk factors associated with its development in a cohort of South African Indians who had been studied 10 years previously. METHODS: This is a report on 563 subjects who participated both at baseline and at the 10-year follow-up study. In the baseline study, 2479 subjects (> 15 years) were studied; using 1985 World Health Organization criteria for glucose tolerance based on 75 g oral glucose tolerance tests (OGTT), the crude prevalence of diabetes mellitus (Diabetes) was 9.8% and of impaired glucose tolerance (IGT) 5.8% (age and sex-adjusted prevalence 13% and 6.9%, respectively). RESULTS: At the 10-year follow-up study, 563 of the subjects who could be traced consented to a repeat OGTT; of these, 91 (16.2%) were classified as Diabetes and 41 (7.3%) as IGT. Of the subjects who did not have diabetes at baseline (n = 517), 49 (9.5%) progressed to diabetes (PTD) and 40 (7.7%) had IGT. The crude cumulative incidence of diabetes was 9.5% (rate of progression 0.95% per annum; incidence density 9.5/1000 person years) with an age and sex-adjusted cumulative incidence of 8.3% (rate of progression 0.95% per annum; incidence density 8.3/1000 person years). Examination of risk factors predictive of subsequent diabetes development was undertaken by analysis of baseline (year 0) variables in the 517 subjects who did not have diabetes at baseline. In multivariate analysis using a logistic regression model, the significant predictive risk factors for future diabetes included 2-h post load plasma glucose (2 PG) (P < 0.0001, odds ratio (OR) 1.7, 95% confidence interval (CI) 1.4-2.1), body mass index (BMI) (P < 0.006, OR 1.1, 95% CI 1.0-1.3) and obesity (P < 0.01, OR 4.6, 95% CI 1.4-14.7). CONCLUSIONS: This long-term study has shown that in South African Indians there is a high incidence of Type 2 diabetes, and in this population significant predictors include higher baseline blood glucose, BMI and obesity.

Asymptomatic hyperparathyroidism caused by a giant parathyroid adenoma.

On routine investigation a 57-year-old woman was found to have primary hyperparathyroidism caused by a giant parathyroid gland. The gland was removed successfully and histological examination proved it to be a parathyroid adenoma.

HLA class II antigens in South African Blacks with type I diabetes.

Type 1 diabetes mellitus is poorly characterised in many African communities, including South Africa, where little is known of the disease epidemiology. This study aimed to identify the HLA class II alleles associated with type 1 diabetes in a group of Zulu subjects in Durban, KwaZulu-Natal by PCR-SSP. The HLA alleles associated with type 1 diabetes included HLA-DQB*0302 (P<0.0001), DRB1*O9 (P<0.0001), DRB1*04 (P=0.002), DRB1*0301 (P=0.003), DQB*02 (P=0.004) and DQA*03 (P=0.035). Estimated haplotypes positively associated with type 1 diabetes included HLA-DRB1 *0301-DQA*0501, DRB1*04-DQA*03, DRB1*04-DQB*0302, DRB1*0301-DQB*0201, DQA*0501-DQB*0201 and DQA*03-DQB*0302. These findings are similar to those reported from Zimbabwe and other populations with type 1 diabetes.

Microvascular complications in South African patients with long-duration diabetes mellitus.

OBJECTIVE: To determine the prevalence of microvascular complications in South African black and Indian patients with long-duration diabetes mellitus (DM). DESIGN: A retrospective analysis was undertaken of clinical records of 219 DM patients (132 black, 87 Indian) with long-duration DM (over 10 years) attending a diabetes clinic in Durban. Data recorded on each subject included demographic details (age, gender, ethnic group, type of diabetes, age of onset and duration of diabetes), presence of retinopathy, markers of nephropathy and biochemical variables. The prevalence of complications and the clinical and biochemical parameters were evaluated for type 1 and type 2 diabetes and for each ethnic group. RESULTS: Of the 219 patients, 47 had type 1 DM (36 blacks, 11 Indians) and 172 were classified as type 2 DM (96 blacks, 76 Indians). The mean age of onset of DM was later in blacks than Indians, both for type 1 (P < 0.05) and type 2 DM (P < 0.01). In patients with type 1 DM, the prevalence of retinopathy was 53.2% (blacks 55.6%, Indians 45.5%), persistent proteinuria was found in 23.4% (blacks 25%, Indians 18.2%) and hypertension in 34%. No ethnic difference was found except for the prevalence of hypertension which was higher in blacks than Indians (41.7% v. 9.1%, P < 0.5). Onset of retinopathy from time of diabetes diagnosis occurred earlier in blacks than Indians (13.0 +/- 4.6 yrs v. 18.0 +/- 4.6 yrs, P < 0.05). For the type 2 DM group, retinopathy was found in 64.5% (black v. Indian 68.8 v. 59.2%) and persistent proteinuria in 25% (black v. Indian 30.2 v. 18.4%). Hypertension was observed in 68% and was more prevalent in blacks (84.4 v. 47.4%, P < 0.01) There was an earlier onset of retinopathy (P < 0.05) and hypertension (P < 0.01) from time of diabetes diagnosis in blacks than Indians. In the type 1 DM group retinopathy was associated with a significantly longer duration of diabetes (P < 0.05) and higher glycated haemoglobin (HbA1) (P < 0.05). For type 2 DM subjects there was a significant association between retinopathy and longer duration of diabetes (P < 0.05) and higher systolic blood pressure (P < 0.05). CONCLUSION: This study has shown that there is a high prevalence of microvascular complications in South African patients with long-duration diabetes mellitus.

Application of the new ADA criteria for the diagnosis of diabetes to population studies in sub-Saharan Africa. American diabetes association.

AIMS: To examine the implications for epidemiological studies of the American Diabetes Association (ADA) recommendation that the fasting blood glucose at a lowered level becomes the main diagnostic test for diabetes on cross-sectional-based data from sub-Saharan Africa. METHODS: Data from 11 surveys conducted in rural, peri-urban and urban Cameroon (n = 1804), South Africa (n = 3799) and Tanzania (n = 10013) which measured fasting (ADA criteria) and 2-h blood glucose concentrations during a standard 75 g OGTT (old WHO criteria) were analysed. RESULTS: The prevalence of diabetes was higher in eight of the 11 surveys when applying the new ADA compared to the old WHO criteria. With the exception of one population (Mara, Tanzania) the absolute difference in prevalence between the two classifications tended to be small (< 2%). There was considerable variation in the categorization of individuals using the ADA and old WHO criteria. The level of agreement between the two ranged from fair to good (Kappa statistic 0.17-0.86). The prevalence of impaired fasting glycaemia (IFG) was lower than that of impaired glucose tolerance (IGT) in 10 of the surveys and the agreement between the two was fair, < or = 0.26 in all the surveys. CONCLUSIONS: Although the use of the new ADA fasting criteria for prevalence surveys is an attractive and practical option, particularly in Africa, further information is required on the characteristics and prognosis of individuals classified as IFG or diabetic by the fasting criteria, prior to wide adoption of the ADA criteria. Ideally measurement of both fasting and two low glucose concentrations should remain the standard for epidemiological studies.