The Public Health Interventions Advisory Committee (PHIAC) took account of a number of factors and issues when developing the recommendations, as follows. Please note: this section does not contain recommendations. (See Recommendations.)
3.1 Evidence suggests that people from black, Asian and other minority ethnic groups are at an equivalent risk of diabetes and other health conditions at a lower body mass index (BMI) than white populations. However, they are not necessarily receiving health promotion advice when their BMI has reached these lower thresholds. PHIAC noted that this may create a significant health inequality. However, the Committee considered that the evidence was insufficient to justify the development of new BMI or waist circumference thresholds to classify whether people in these groups in England are overweight or obese.
3.2 PHIAC noted that evidence of 'equivalence of risk' and the need to intervene at a lower BMI may not be the same as evidence on 'equivalence of response' to interventions. In other words, people from black, Asian and other minority ethnic groups may have the same risk of mortality and diabetes at a lower BMI, compared to white populations. However, they may not respond in the same way to behaviour-change interventions as white populations. Or, if they do lose weight, they may not gain the same benefit as someone who is white. The evidence considered did not allow for a prediction of response to behavioural interventions, as this was beyond the scope of the guidance.
3.3 The lack of precise BMI and waist thresholds, whereby the risk of a range of health conditions could be identified for black, Asian and minority ethnic groups, could result in widening inequalities in health.
3.4 PHIAC noted that there are recognised differences in terms of health outcomes within ethnic groups and it is important to note that these groups are not homogeneous (Nazroo 2004).
3.5 No single threshold on BMI and waist circumference for all minority ethnic groups, across a range of conditions, was found in the evidence. However, the evidence did clearly show that black and Asian populations suffer from adverse health outcomes at a lower BMI than people of white ethnicity – although the precise cut-off points were uncertain and most evidence related to diabetes.
3.6 The evidence supports use of the World Health Organization's (WHO's) public health action points for intervening to prevent diabetes. It also supports the recommendations made in Preventing type 2 diabetes: risk identification and interventions for individuals at high risk (NICE public health guidance 38) for Asian populations. In addition, the evidence indicated that the threshold range for Asian populations may be extended to black populations. However, it was equivocal (or non-existent) in relation to the question of where to set BMI and waist circumference thresholds as a marker of general health risks or mortality for black, Asian and other minority ethnic groups.
3.7 NICE's Obesity: the prevention, identification, assessment and management of overweight and obesity in adults and children (NICE clinical guideline 43), published in 2006, did not make a recommendation on specific BMI and waist circumference cut-off points for different minority ethnic groups. However, it did note that: '…some other population groups, such as Asians and older people, have comorbidity risk factors that would be of concern at different BMIs (lower for Asian adults and higher for older people).' It also advised that healthcare professionals should '…use clinical judgement when considering risk factors in these groups, even in people not classified as overweight or obese using the classification in recommendation 18.104.22.168.' PHIAC considered that the evidence accumulated since 2006 is still insufficient to make specific recommendations about BMI and waist circumference thresholds for classifying whether a person in these groups is overweight or obese.
3.8 Some studies included in the evidence review used self-reported measures of waist circumference, BMI and health outcomes (for example, on diabetes). PHIAC noted that this may have introduced measurement error, and bias.
3.9 The relationship between ethnicity and obesity is complex and not all studies were adjusted for the same potentially confounding factors.
3.10 Any estimate of equivalence will include a degree of uncertainty, irrespective of the method used. Some of the equivalence thresholds discussed by PHIAC were particularly likely to be imprecise, as they were derived post-hoc by the evidence reviewers without the original data. This was done by using figures found in the published literature and drawing a horizontal line that intersects the incidence or prevalence rates to estimate risk equivalence between white and black, South Asian or Chinese populations.
3.11 PHIAC recognised that ongoing UK studies may provide published evidence on BMI thresholds in the future. These include the 'Southall and Brent revisited' (SABRE) cohort and the Leicester cohort of the 'Anglo-Danish-Dutch study of intensive treatment in people with screen detected diabetes in primary care' (ADDITION). This evidence was available as non-peer reviewed expert testimony at the post-consultation PHIAC meeting in March 2013. It was undergoing peer review, in preparation for publication as an academic paper, when this guidance was published.
3.12 Being classified as obese at a lower BMI or waist circumference threshold has a number of potential disadvantages. For example, someone might feel labelled, stigmatised or may, in some other way, be harmed psychologically. Any potential harm may be made worse if they gain little benefit from being offered a lifestyle intervention at a lower BMI threshold. However, PHIAC noted that if people are at equivalent risk at a lower BMI, then the benefits of offering behavioural support at a lower threshold are likely to outweigh any ill effects.
3.13 Patient notes do not always include BMI or waist circumference measures. Waist circumference, in particular, is rarely noted by GPs. In addition, information on ethnicity is often not recorded.
3.14 Health professionals may be unaware of the disproportionate risks and burden of disease that black, Asian or other minority ethnic groups face when classified as overweight or obese using BMI thresholds that may be more appropriate for white European populations.
3.15 Other approaches to anthropometric measurement, such as waist-to-hip and waist-to-height ratio, were not assessed. This should not be taken as a judgement on whether or not these approaches are effective.
3.16 PHIAC did not consider evidence on the effectiveness or cost effectiveness of intervening at different BMI and waist circumference thresholds for different black, Asian and other minority ethnic groups. This was not part of the scope of the guidance.