IntroductionBMI is an established way of classifying the degree of excess weight in an individual. Nearly three decades ago in 1993, a WHO committee of experts proposed BMI cutoffs of 25·0–29·9 kg/m2 for overweight grade 1, 30·0–39·9 kg/m2 for overweight grade 2 (now termed obesity class I), and 40·0 kg/m2 or more for overweight grade 3 (now termed obesity class III).1WHO
Physical status: the use and interpretation of anthropometry. The suggested BMI cutoff now used to define obesity (≥30 kg/m2) was developed from observational studies in Europe and the USA of exclusively White populations and based on the association between BMI and mortality. Subsequently, there has been increasing evidence of a high prevalence of type 2 diabetes among Asian populations at a lower BMI than in White populations.2Universal cut-off BMI points for obesity are not appropriate., 3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. In response to these emerging data, WHO recommended lowering the BMI cutoffs for defining obesity in south Asian populations to optimise the identification of cardiometabolic risk in this group.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Originating from a WHO expert consultation in 2004,3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. WHO, and, subsequently, the National Institute for Health and Care Excellence (NICE), recommended a BMI cutoff of 27·5 kg/m2 be used for south Asian and Chinese populations to trigger the implementation of lifestyle interventions.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies., 4National Institute for Health and Care Excellence
BMI: preventing ill health and premature death in black, Asian and other minority ethnic groups. Public health guideline [PH46]. The expert consultation recalculated BMI cutoffs based on the measurement of percentage body fat, which is typically higher in Asian people than in White people, from studies done in China, Hong Kong, Indonesia, Japan, Singapore, and Thailand.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Despite the importance of identifying BMI cutoffs for obesity at which adverse outcomes, such as type 2 diabetes, are more likely to occur and producing clinically relevant guidelines for patient care, WHO made recommendations with no or sparse data on the association of BMI with type 2 diabetes and without data on Black, south Asian, and Arab populations.Type 2 diabetes can be prevented or delayed through dietary change, physical activity, or the use of metformin.5Haw JS Galaviz KI Straus AN et al.Long-term sustainability of diabetes prevention approaches: a systematic review and meta-analysis of randomized clinical trials. The early use of other antihyperglycaemic therapies reduces the risk of long-term complications from type 2 diabetes via improved glycaemic control.6Marso SP Daniels GH Brown-Frandsen K et al.Liraglutide and cardiovascular outcomes in type 2 diabetes., 7Wiviott SD Raz I Bonaca MP et al.Dapagliflozin and cardiovascular outcomes in type 2 diabetes., 8Perkovic V Jardine MJ Neal B et al.Canagliflozin and renal outcomes in type 2 diabetes and nephropathy., 9UK Prospective Diabetes Study (UKPDS) Group
Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). However, these benefits cannot be fully realised if current WHO and NICE recommendations for obesity under-recognise the risk of developing type 2 diabetes in minority ethnic populations.Research in context

Evidence before this study

WHO and the National Institute for Health and Care Excellence (NICE) in England both recommend a BMI cutoff of 27·5 kg/m2 to trigger action to reduce the risk of obesity-related conditions, such as type 2 diabetes, in south Asian and Chinese populations. This recommendation is based on a sparse evidence base and therefore might be inappropriate for some minority ethnic groups. Previous studies have attempted to identify BMI cutoffs for obesity in multi-ethnic populations by use of data on type 2 diabetes prevalence or a surrogate marker, small sample sizes, and self-reported disease status, including relatively few minority ethnic groups. Because type 2 diabetes can be delayed or prevented through dietary change, physical activity, and the early use of antihyperglycaemic therapy, it is important to establish BMI cutoffs for obesity in relation to the risk of type 2 diabetes among adults from minority ethnic populations in England that equate to those developed in White populations.

Added value of this study

In a comprehensive analysis, we define BMI cutoffs for obesity based on the risk of developing type 2 diabetes in minority ethnic adults equivalent to the BMI cutoff for obesity of 30·0 kg/m2 set for White populations. To our knowledge, this study is the first to provide BMI cutoffs for obesity for Arab populations and Black and south Asian ethnic subgroups. We also highlight the value of routine electronic health records and the use of large, linked datasets to provide precise ethnicity-specific BMI cutoffs for obesity. In our study, we included 1 472 819 people aged 18 years or older registered with a general practitioner practice in England at any point between 1990 and 2018 (1 333 816 were White, 75 956 were south Asian, 49 349 were Black, 10 934 were Chinese, and 2764 were Arab). For an equivalent age-adjusted and sex-adjusted incidence of type 2 diabetes at a BMI of 30·0 kg/m2 in White populations, we found lower BMI cutoffs for south Asian (23·9 kg/m2), Black (28·1 kg/m2), Chinese (26·9 kg/m2), and Arab (26·6 kg/m2) populations.

Implications of all the available evidence

By contrast to WHO expert consultation recommendations and NICE guidelines, our study shows that Black Caribbean, south Asian, Chinese, and Arab populations living in England had an equivalent risk of type 2 diabetes at substantially lower BMI values than the current BMI cutoffs for obesity. Our findings should guide revisions of current ethnicity-specific BMI cutoffs to trigger action to reduce the risk of developing type 2 diabetes and equalise opportunities for the increased prevention and early diagnosis of type 2 diabetes.

Several further attempts have been made to establish ethnicity-specific BMI cutoffs to identify obesity in relation to type 2 diabetes risk in multi-ethnic populations based in the UK and North America.10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants., 12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population., 13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. Such attempts had several limitations: the studies used prevalence data for type 2 diabetes11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants. or a surrogate marker,12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population. lacked precision because of small ethnic group sizes,10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants., 12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population., 13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. and did not examine particular minority ethnic groups.10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants., 12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population., 13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. To address these challenges, we used a large-scale, longitudinal database of linked primary and secondary care electronic health records from a representative sample of the population in England to identify BMI cutoffs for obesity based on the risk of developing type 2 diabetes among adults from Black, south Asian, Chinese, and Arab populations in England equivalent to the BMI obesity-related cutoff of 30·0 kg/m2 established in White populations.ResultsFrom a total of 2 249 438 individuals aged 18 years or older with no previous diagnosis of diabetes and with any follow-up, recruited between Sept 1, 1990, and Dec 1, 2018, 1 472 819 were included in the study (figure 1). 1 333 816 (90·6%) participants were White, 75 956 (5·2%) were south Asian, 49 349 (3·4%) were Black, 10 934 (0·7%) were Chinese, and 2764 (0·2%) were Arab (table; appendix p 20). More women than men were included in the study, and, at baseline, the mean BMI was lowest among the Chinese group than among the other ethnic groups (table).

TableBaseline characteristics of the study population by ethnicity

Data are mean (SD) or n (%).

After a median follow-up of 6·5 years (IQR 3·2–11·2), 97 823 (6·6%) of the 1 472 819 individuals in our study were diagnosed with type 2 diabetes. Of the 97 823 diagnosed participants, 89 287 (91·3%) were White, 5632 (5·8%) were south Asian, 2444 (2·5%) were Black, 317 (0·3%) were Chinese, and 143 (0·1%) were Arab. The median age at diagnosis of type 2 diabetes was 67 years (IQR 57–76) in White individuals, 55 years (45–65) in south Asian individuals, 54 years (47–65) in Black individuals, 60 years (52–68) in Chinese individuals, and 56 years (47–64) in Arab individuals.

For the equivalent age-adjusted and sex-adjusted incidence of type 2 diabetes at a BMI of 30·0 kg/m2 in White populations, the BMI cutoffs were 23·9 kg/m2 (95% CI 23·6–24·0) in south Asian populations, 28·1 kg/m2 (28·0–28·4) in Black populations, 26·9 kg/m2 (26·7–27·2) in Chinese populations, and 26·6 kg/m2 (26·5–27·0) in Arab populations (figure 2). Further adjustment for self-reported smoking status and socioeconomic position did not substantially change the estimated ethnicity-specific BMI cutoffs for obesity (appendix p 21). The BMI cutoffs for Black ethnic subgroups and south Asian ethnic subgroups equivalent to a BMI of 30·0 kg/m2 for White populations, related to the age-adjusted and sex-adjusted incidence of type 2 diabetes, can be found in figure 3.

Figure 2Age-adjusted and sex-adjusted BMI cutoffs in minority ethnic populations in England equivalent to a BMI cutoff of 30·0 kg/m2 in White populations in relation to type 2 diabetes incidence

Show full caption

The incidence of type 2 diabetes for a BMI of 30·0 kg/m2 in the White population can be read off the graph at the intersection of the grey horizontal line and the fitted line for the White population.

Figure 3Age-adjusted and sex-adjusted BMI cutoffs in minority ethnic subgroups in England equivalent to a BMI cutoff of 30·0 kg/m2 in White populations in relation to type 2 diabetes incidence

Show full caption

The incidence of type 2 diabetes for a BMI of 30·0 kg/m2 in the White population can be read off the graph at the intersection of the grey horizontal line and the fitted line for the White population.

For the equivalent age-adjusted and sex-adjusted incidence of type 2 diabetes at a BMI of 25·0 kg/m2 in White populations, the BMI cutoffs were 19·2 kg/m2 (95% CI 18·9–19·3) in south Asian populations, 23·4 kg/m2 (23·2–23·6) in Black populations, 22·2 kg/m2 (22·0–22·4) in Chinese populations, and 22·1 kg/m2 (21·8–22·0) in Arab populations (figure 4).

Figure 4Age-adjusted and sex-adjusted BMI cutoffs in minority ethnic populations in England equivalent to a BMI cutoff of 25·0 kg/m2 in White populations in relation to type 2 diabetes incidence

Show full caption

The incidence of type 2 diabetes for a BMI of 25·0 kg/m2 in the White population can be read off the graph at the intersection of the grey horizontal line and the fitted line for the White population.

Discussion

Using electronic health records from approximately 1·5 million individuals, of whom 97 823 were diagnosed with type 2 diabetes during a median follow-up of 6·5 years, we provide new BMI thresholds for obesity to trigger action to reduce the risk of developing type 2 diabetes in Black, south Asian, Chinese, and Arab populations living in England. Our data address the ongoing debate around the interpretation of recommended BMI cutoffs for identifying obesity in minority ethnic populations. For an equivalent age-adjusted and sex-adjusted incidence of type 2 diabetes at a BMI of 30·0 kg/m2 in White populations, we found lower BMI cutoffs for south Asian (23·9 kg/m2), Black (28·1 kg/m2), Chinese (26·9 kg/m2), and Arab (26·6 kg/m2) populations.

Obesity, defined as a BMI of 30·0 kg/m2 or more, is a widely used measure and an important risk factor for the development of type 2 diabetes.21Chatterjee S Khunti K Davies MJ However, the appropriateness of this BMI cutoff in non-White minority ethnic populations is contentious, and remains a subject of debate because of important limitations in the evidence base.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies., 4National Institute for Health and Care Excellence
BMI: preventing ill health and premature death in black, Asian and other minority ethnic groups. Public health guideline [PH46]., 22Hsu WC Araneta MR Kanaya AM Chiang JL Fujimoto W BMI cut points to identify at-risk Asian Americans for type 2 diabetes screening., 23Ethnic-specific criteria for classification of body mass index: a perspective for Asian Indians and American Diabetes Association position statement. Previously reported studies that attempted to identify BMI cutoffs for obesity in multi-ethnic populations relied on prevalence data,12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population. in which BMI and type 2 diabetes status were ascertained at the same timepoint, lacked precision because of small sample sizes,10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population., 13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. used surrogate markers12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population. and self-reported data,10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants. and included relatively few ethnic groups.13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom.WHO and NICE both recommend a BMI cutoff of 27·5 kg/m2 to define obesity in south Asian and Chinese populations to trigger lifestyle interventions.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies., 4National Institute for Health and Care Excellence
BMI: preventing ill health and premature death in black, Asian and other minority ethnic groups. Public health guideline [PH46]. NICE also suggest that this lower BMI threshold should be used to trigger action to prevent type 2 diabetes among Black populations.24National Institute for Health and Care Excellence
Obesity: identification, assessment and management. Clinical guideline [CG189]. Our study clearly showed that, compared with the risk of developing type 2 diabetes at a BMI of 30·0 kg/m2 in White populations, the equivalent risk among south Asian individuals occurred at a BMI of 23·9 kg/m2, a cutoff much lower than the recommended ethnicity-specific cutoff of 27·5 kg/m2. Our findings are consistent with previous studies in suggesting that the cutoffs currently recommended by WHO and NICE should be reduced when applied to non-White populations.10Chiu M Austin PC Manuel DG Shah BR Tu JV Deriving ethnic-specific BMI cutoff points for assessing diabetes risk., 11Ntuk UE Gill JM Mackay DF Sattar N Pell JP Ethnic-specific obesity cutoffs for diabetes risk: cross-sectional study of 490 288 UK biobank participants., 12Razak F Anand SS Shannon H et al.Defining obesity cut points in a multiethnic population., 13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. For example, in the SABRE (Southall and Brent Revisited) cohort study13Tillin T Sattar N Godsland IF Hughes AD Chaturvedi N Forouhi NG Ethnicity-specific obesity cut-points in the development of type 2 diabetes—a prospective study including three ethnic groups in the United Kingdom. of Europeans (n=1356), south Asians (n=842), and African-Caribbeans (n=335) in north and west London, UK, aged 40–69 years at baseline (recruited between 1988 and 1991) and followed up for a median of 19 years, age-adjusted and sex-adjusted BMI cutoffs for obesity were 25·2 kg/m2 for south Asians and 27·2 kg/m2 for African-Caribbeans. We found that the incidences of type 2 diabetes among south Asian subpopulations (ie, Indian, Pakistani, Bangladeshi, Nepali, Sri Lankan, and Tamil) were equivalent to that in the White population at consistently much lower values of BMI. However, when examining Black ethnic subgroups (ie, Black Africans, Black Caribbean, Black British, and other Black people), we found that the incidences of type 2 diabetes were equivalent to that in the White population at lower BMI values only for Black Caribbean individuals and Black people of other ethnic origins. We also found that BMI cutoffs for overweight based on the risk of type 2 diabetes were lower for south Asian, Black, Chinese, and Arab populations than for White populations (25·0 kg/m2), suggesting that the recommended BMI cutoff for overweight to trigger action to reduce the risk of type 2 diabetes should also be lowered in these groups. Whether lower BMI cutoffs in non-White populations are due to differences in body composition, biochemical characteristics, lifestyle factors (eg, physical activity or diet), the genetic architecture of type 2 diabetes, or lifestyle–gene interactions remains unclear.25Ethnicity and type 2 diabetes in the UK., 26Eastwood SV Tillin T Dehbi HM et al.Ethnic differences in associations between fat deposition and incident diabetes and underlying mechanisms: the SABRE study., 27Genomics, type 2 diabetes, and obesity. Future studies that examine the relative contributions of these mechanisms to the development of type 2 diabetes might help to explain our study findings.A limitation of our study is that, even though individuals registered in the CPRD are representative of the general adult population in the UK,14Herrett E Gallagher AM Bhaskaran K et al.Data resource profile: Clinical Practice Research Datalink (CPRD). individuals with a recorded BMI measurement might not necessarily be representative of the general UK population. BMI data, if not recorded as part of registration with a general practitioner, tends to be recorded opportunistically (ie, when the patient is using health-care services for other reasons or when a BMI measurement is of direct clinical importance). We reduced this possibility by only using the first BMI value recorded from the registration date (these values would have probably been recorded for administrative and not health reasons). Our findings of lower BMI cutoffs for obesity in minority ethnic populations compared with White populations living in England apply only to the risk of developing type 2 diabetes, and might not apply to other endpoints, such as cardiovascular disease or all-cause mortality. Type 2 diabetes was chosen as the outcome of interest because it is the most specific obesity-related complication and a chronic, progressive disease with considerable health and socioeconomic costs.21Chatterjee S Khunti K Davies MJ BMI is a simple, inexpensive surrogate measure of body fat used in primary and secondary care, and is the subject of national and international guidelines on assessing adiposity.3WHO expert consultation
Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies., 4National Inst