A guide to using height, weight, body mass index and other body measurements in carrying out nutritional assessments
Lin Perry, PhD, MSc, RGN, RNT is senior research fellow, Newcastle Institute of Public Health, University of Newcastle, New South Wales, Australia
Body measurements have been used as nutritional indices for many years. Some, such as height, are associated with previous nutritional conditions, while others reflect more recent nutritional status. Some measurements can distinguish between fat and fat-free mass, providing separate information about energy and muscle protein stores.
This is most often used as a nutritional index with children. Stunted growth is defined as height measurement two standard deviations below median height for age. It is also required to calculate body mass index (BMI).
However, height measurement is of limited use for a number of reasons. Accurate measurement is not always possible as it requires the subject to stand erect under a stadiometer, which may not be available, while many patients are unable to stand. In addition, it is not clear how measurements should be interpreted in people who have experienced height loss due to conditions such as kyphosis (curvature of the upper spine) or scoliosis (side-to-side curvature of the spine). Recalled or reported height may be used but there is a risk of incurring wide margins of error.
Alternative methods provide height measurements for BMI calculation. Most have error margins of around 4cm but may be useful and acceptable with frail people.
This is measured with the arm outstretched straight, palm forwards, from the base of the middle/ring fingers to the sternal notch using a non-stretch tape measure. A calculation (Box 1) provides a ‘height’ value.
This is not an easy measurement to take with people who are confined to bed, have upper limb disabilities or chest and back deformities, such as kyphosis or scoliosis. In this situation, other approaches such as ulnar length are easier and preferable.
This is measured either lying supine with the knee flexed at 90 using knee-height calipers (Chumlea et al, 1985) or in the seated position using a stadiometer (Haboubi et al 1990).
Calculations for height approximations are described in Box 2 and nomograms (calculating device) are available (Haboubi et al, 1990; Chumlea et al, 1985).
An alternative approach is to measure knee height with the person sitting on a chair, without footwear, with their knee at a right angle. The left leg should be used for measurement if possible.
Measurements should be taken with a tape measure that has a ring or extra length beyond the zero point, so that the 0 reading can be underneath the third or fourth fingers.
The hand of the person taking the measurement should be positioned flat across the thigh of the person being measured, approximately 4cm behind the front of the knee;
The tape measure is extended straight down the side of the leg, across the bony prominence at the ankle (lateral malleolus) to the base of the heel.
This is measured by asking the person to bend an arm up diagonally (left arm preferably), so that the palm is flat across chest, fingers pointing to the opposite shoulder;
The length is measured in centimetres to the nearest 0.5cm between the point of the elbow (olecranon) and the mid-point of the prominent bone of the wrist (styloid process) (Todorovic et al, 2003);
Ulnar length is one of the options suggested for use in the calculation of Body Mass Index as part of the Malnutrition Universal Screening Tool.
Malnutrition Universal Screening Tool includes a table of the equivalent height for ulnar length measurements for men and women under and over the age of 65 years (Todorovic et al, 2003).
Weight is extensively used as a nutritional measurement both for assessment and monitoring. It has a number of well-recognised limitations:
Accurate measurement requires regularly calibrated and maintained scales capable of accommodating patients who are confined to bed;
Clothing, fluid balance including diurnal or monthly cycles, oedema and dehydration, surgery and amputations, tumour load and alterations of drug prescriptions can all influence weight;
Weight by itself is difficult to interpret although change over time may be more informative. Percentage weight change can be calculated (Box 3). Significant weight loss has been suggested as >5% loss in one month or >10% in six months although 5-10% loss within 3-6 months may indicate nutritional risk (Todorovic et al, 2003). Loss of >5% over one year has been linked with increased mortality (Wallace et al, 1995).
Current weight can also be compared to usual weight, if this is known or against tables of ideal or reference weights. The calculation is shown in Box 4. However, many of these ideal and reference weights are dated and American in origin and it is not clear how relevant they are for people in the UK.
Body mass index
Weight alone takes no account of frame size; this is achieved in combination with height. At least four versions of BMI are available but Quetelet’s Index is most often used for adults; Ponderal’s Index for children (Lee et al, 1981).
Based on mortality and morbidity tables, adults are often considered underweight at <18.5kg/m2 although in older people a higher cut-off (possibly 23kg/m2) may be indicated, due to relative changes in body composition with ageing (Beck and Ovesen, 1998). Obesity is defined as values above 30kg/m2. BMI does not distinguish fat from muscle.
There are standard procedures for skinfold thickness measurements, which can be taken at a number of sites (Lee and Nieman, 1996). These include the abdomen, thigh, chest and below the scapula, although the triceps is the most accessible.
Measurements may be used to predict proportions of body fat using several sites but used in this way relies on various assumptions about tissue compressibility and body fat distribution (Lee and Nieman, 1996).
More frequently, single-site measurements are used. These may be compared to tables of population percentile values although the same limitations apply as weight tables, particularly with older people. Most often, repeated triceps skinfold measurements are used for medium or long-term monitoring of nutritional progress.
Calibrated callipers and a tape measure are required and staff need training and regular practice to achieve and maintain accuracy.
Mid arm circumference measurement includes fat, muscle and bone. An equation allows calculation of muscle protein stores (Box 5). This produces results within +/- 8% of computerised tomogram (CT) measurements of arm muscle (Heymsfield et al 1982), but may still result in an over-estimation in obese people. Like skinfold thickness, it is not sensitive enough to indicate short-term changes.
Measurements should be taken in a standing or sitting position.
Clothing should be removed from the left arm if possible, and the top of the shoulder (the acromion) and the point of the elbow (olecranon process) located.
The distance between these points should be measured, and the mid-point identified and marked on the arm. The person should then be asked to let the arm hang down loose. The circumference of the arm can then be measured at the marked mid-point, with the tape measure fitting comfortably but not tight round the arm (Todorovic et al 2003).
If it is not possible for the person to stand or sit, both arm circumference and triceps skinfold thickness can be taken with subjects supine or lying on their side, but the error margin will be higher.
If neither weight nor height can be measured, arm circumference can also be used to provide a broad estimate of BMI. If the measured mid-arm circumference is less than 23.5cm, BMI is likely to be less than 20kg/m2 and the person is likely to be underweight. Conversely, if arm circumference is more than 32.0cm, BMI is likely to be more than 30kg/m2, and the person is likely to be obese (Todorovic et al, 2003). Accuracy may be increased if two measurements are taken at each assessment and the average of both is calculated and recorded.
This is a rough indication only, but can also be used to provide a guide to changes over time. Repeated measurements should be taken over a period of time and the trend reviewed. If there is a change of at least 10%, it is likely that the person’s weight and BMI will have changed by a similar amount (Todorovic et al, 2003).
Other measurements are more demanding and seldom used in clinical practice for example, densitometry. Bio-electrical impedance, used to calculate body fat in many gyms is heavily influenced by hydration so is less useful in acutely ill patients and should not be undertaken in anyone with an implantable cardiac pacemaker.
Anthropometric measurements may help with making nutritional judgements but the circumstances of individual patients and the accuracy and repeatability of the measurements must be born in mind when a patient is being reviewed.
Bassey, E.J. (1986) Demi-span as a measure of skeletal size. Annals of Human Biology; 13: 5, 499-502.
Beck, A.M., Ovesen, L. (1998) At which Body Mass Index and degree of weight loss should hospitalised elderly patients be considered at nutritional risk? Clinical Nutrition; 17: 5, 195-198.
Chumlea, W.C. et al (1985) Estimating stature from knee height for persons 60 to 90 years of age. Journal of the American Geriatrics Society; 33: 2, 116-120.
Haboubi, N.Y. et al (1990) Measurement of height in the elderly. Journal of the American Geriatrics Society; 38: 9, 1008-1010.
Heymsfield, S.B. (1982) Anthropometric measurement of muscle mass: revised equations for calculating bone-free arm muscle area. American Journal of Clinical Nutrition; 36: 4, 680-690.
Lee, J. et al (1981) Relative merits of the weight-corrected-for-height indices. American Journal of Clinical Nutrition: 34; 11, 2521-2529.
Lee, R.D., Nieman, D.C. (1996) Nutritional Assessment. St Louis: Mosby
Todorovic, V. et al (eds) on behalf of the Malnutrition Advisory Group (2003) The ‘MUST’ Explanatory Booklet. A Guide to the ‘Malnutrition Universal Screening Tool’ (‘MUST’) for Adults. British Association for Parenteral and Enteral Nutrition.
Wallace, J.I. et al (1995) Involuntary weight loss in older outpatients: incidence and clinical significance. Journal of the American Geriatrics Society; 43: 4, 329-337.