BMI is an appropriate measure in spite of limitations

BMI as a measure of population prevalence of overweight and obesity has certain limitations. Although generally related to body fat, it does not measure fat itself. BMI does not take into account skeletal size, amount of body water or muscle mass and it is not gender specific. Nevertheless, on a population level BMI is considered to be the most appropriate measure. It is easily calculated on the basis of standard measures that generally show little interobserver variation. Evaluation of BMI is inexpensive, it requires minimal training, and results are immediately available.

Self reported data tend to underestimate overweight

BMI can be monitored by means of questionnaires (self reported) or actual measurements of height and weight. Since people with obesity tend to underreport their bodyweight, self reported data tend to underestimate the scale of the overweight problem. Furthermore, errors in self reported height and weight may vary with age and overweight status (Kuczmarski et al., 2001).

Obesity in children differs from obesity in adults

Obesity in children is different from obesity in adults in some important respects. The main difference is that children and adolescents are growing. So for example, during puberty, a child’s weight will double and its height will increase by 20%. Simple measures of obesity such as the body mass index (BMI) cannot be used directly because it underestimates the degree of overweight in short children and overestimates overweight in tall children (Asp et al., 2002). The IOTF’s international standard for analysing childhood overweight and obesity data has now been widely adopted (IOTF, 2005b; Cole et al., 2000). It provides growth curves which relate cut-off points for different age groups to the adult categories for overweight and obesity (BMI 25-29.9 and BMI≥30 respectively; see International cut off points for BMI).

Ethnic differences in percentage of fat and fat distribution

Percentage of body fat and fat distribution vary between different population groups. Apart from age and sex, ethnicity and physical activity level affect the percentage of body fat. For example, endurance runners have less body fat than swimmers. People living near the North Pole generally have more body fat than people living in more moderate temperature zones. Ethnic differences in fat proportion and corresponding risk levels might also stem from differences in relative leg lengths and/or in frame size (Deurenberg et al., 2002).

Debate on different BMI cut-off points for different ethnic groups

Due to differences across populations, there has been a debate on whether different BMI cut-off points for different ethnic groups should be developed. For example, a BMI of 27.5 in an Asian person may be associated with comparable morbidities to those seen in a Caucasian person with a BMI of 30. However, available data indicate amplified risks for Asians below conventional overweight or obesity markers. Therefore, an action point of BMI>23 was proposed by a WHO expert group, which nevertheless agreed that existing international classifications should be retained. The consultation identified potential public health action points (23.0, 27.5, 32.5, and 37.5 kg/m²) along the continuum of BMI, and proposed methods by which countries could make decisions about the definitions of increased risk for their population (WHO Expert Consultation, 2004).