Arm circumference as an index of protein-energy malnutrition in six- to eleven-month old rural Tanzanian children

The use of arm circumference (AC) as an inexpensive indicator of protein-energy malnutrition (PEM) has been largely confined to children 12-59 months old. In this age range, AC has been shown to be relatively age-independent and highly correlated with traditional weight-for-age (W/A) indices. We analysed AC for these two properties in this anthropometric survey of rural Tanzanian children 6-11 months (n = 85) and 12-59 months (n =123) old. We compared W/A to NCHS 50th percentile standards and AC to Wolanski standards. AC cut-offs tested included a 13.5-cm mark for both age groups and 12.5 cm for those 6-11 months. The correlation between AC and W/A was 0.8214 in the 6-11-month group and 0.6771 for the 12-59-month age range. Using a W/A standard and a 13.5-cm cut off, AC had a sensitivity of 88 per cent and specificity of 57 per cent in the 6-11-month group, and respective values of 69 and 76 per cent in the 12-59-month group. In those 6-11 months old, sensitivity fell to 79 per cent and specificity rose to 95 per cent with a 12.5-cm cut off. In Wolanski's AC data, AC increased 10 per cent from 6 to 11 months (1.7 per cent/rath) and 8 per cent from 12 to 59 months (0.17 per cent/mth). Thus, despite the greater rate of AC increase in the first year of life, the absolute increase is small enough that, for the vulnerable 6-11-month age group in this region of Tanzania, a 12.5-cm cut off identifies PEM with high specificity and appropriate sensitivity to promote early intervention at the village level.