BMI calculation in older people: The effect of using direct and surrogate measures of height in a community-based setting.

Background & Aims: There is currently no consensus on which measure of height should be used in older people's body mass index (BMI) calculation. Most estimates of nutritional status include a measurement of body weight and height which should be reliable and accurate, however at present several different methods are used interchangeably. BMI, a key marker in malnutrition assessment, does not reflect age-related changes in height or changes in body composition such as loss of muscle mass or presence of oedema. The aim of this pilot study was to assess how the use of direct and surrogate measures of height impacts on BMI calculation in people aged ≥75 years.

Methods: A cross-sectional study of 64 free-living older people (75-96 yrs) quantified height by two direct measurements, current height (H), and self-report (H) and surrogate equations using knee height (H) and ulna length (H). BMI calculated from current height measurement (BMI) was compared with BMI calculated using self-reported height (BMI) and height estimated from surrogate equations for knee height (BMI) and ulna length (BMI).

Results: Median difference of BMI-BMI was 2.31 kg/m. BMI gave the closest correlation to BMI. The percentage of study participants identified at increased risk of under-nutrition (BMI < 20 kg/m) varied depending on which measure of height was used to calculate BMI; from 5% (BMI), 7.8% (BMI), 12.5% (BMI), to 14% (BMI) respectively.

Conclusions: The results of this pilot study in a relatively healthy sample of older people suggest that interchangeable use of current and reported height in people ≥75 years can introduce substantial significant systematic error. This discrepancy could impact nutritional assessment of older people in poor health and lead to misclassification during nutritional screening if other visual and clinical clues are not taken into account. This could result in long-term clinical and cost implications if individuals who need nutrition support are not correctly identified. A consensus is required on which method should be used to quantify height in older people to improve accuracy of nutritional assessment and clinical care.