amount of these supplements consumed in the United States and Canada are unavailable.

Because there is no information from national surveys on choline intakes or on supplement usage, the risk of adverse effects within the United States or Canada can not be characterized.

Sufficient human data are not available for determining whether choline is essential in the human diet, how much is required if it is essential, and the public health impact of poor choline nutriture. For this reason, research that could provide such human data is assigned the highest priority:

• Examination of the effects of the use of graded levels of dietary intake of choline on parameters of health. This would include assessing plasma and tissue choline compounds and metabolites; plasma cholesterol and homocysteine concentrations; erythrocyte folate; and liver, renal, brain, and other organ function. To facilitate this process, food composition data are needed for choline, phosphocholine, glycerophosphocholine, sphingomyelin, phosphatidylcholine, and betaine and the analytic sensitivity and specificity of methods for analysis of food composition need to be validated.

• Human studies on interrelationships among requirements for choline, methionine, folate, vitamin B₆, and vitamin B₁₂ to compare the homocysteine-lowering effects of combinations of these nutrients.

Two additional topics also merit attention:

• The relative effectiveness of different choline-containing compounds in the diet in promoting health and determination of the sparing effect of endogenous synthesis of choline. It will be important to conduct studies on the bioavailability of choline and choline compounds and on the rate of de novo synthesis of choline in vivo.

• Studies using increasing levels of dietary intake designed to assess toxicity for all organ systems, including heart, liver, brain and