Hazards of lumbar puncture.

SIR,-Mr J C Hall's comments about examining reproducibility and using sensitivityspecificity analysis are perfectly valid, but in a short report there is a limit to the amount of information that can be included. We have in fact carried out studies concerned with both these aspects (unpublished observations), and midarm circumference seems more satisfactory than skinfold thickness measurements and derived measurements such as midarm muscle circumference. We concentrated on outpatients rather than inpatients because we believe that it is in this group particularly that malnutrition is overlooked. Inpatients are often subjected to a variety of laboratory investigations, and nutritional deficiencies are likely to be identified without the need for a simple screening test. There has been recent concern about the appropriateness of reference standards derived by Jelliffe in 1966. 1 2These standards, however, are still used extensively in nutritional assessments, and they have the advantage of simplicity, long term use, and easy commitment to memory. Any anthropometric reference standard should be related to functional outcome-that is, morbidity and mortalityand there is no evidence as yet that the centiles published by Frisancho3 are in any way more valid than Jelliffe's standards. Malnutrition is difficult to define, and "gold standards" would be hard to find. We have simply found that in Crohn's disease patients with thin arms are more likely to have nutritional deficiencies than those with fat arms, and this deserves further investigation.