Better Sleep Through Carbs?

Those who have trouble falling asleep at night might consider reaching for a bowl of jasmine rice instead of a sleeping pill, based on results of a recent Australian study (High-glycemic-index carbohydrate meals shorten sleep onset, Afaghi A, OConnor H, and Chow CM, The American Journal of Clinical Nutrition, Feb. 2007; 85:426-430).

Because carbohydrate intake increases the plasma concentration of tryptophan, a precursor of serotonin and a sleep-inducing compound, University of Sydney researchers decided to explore the effect of glycemic index (GI) and the timing of meals on sleep. The scientists tested 12 healthy men, aged 18 to 35 years, with no sleep problems. After fasting for five hours they ate a meal of rice with steamed vegetables in tomato puree, with one of two types of rice and different mealtimes. All the participants tested each of three variables: One test meal was made with high-GI (109) jasmine rice and eaten either one hour or four hours before bedtime. Another meal included lower-GI (50) long-grain rice and was also eaten four hours before bedtime.

The two types of rice contain similar levels of carbohydrates, 35% to 36%, but differ in starch composition. Although jasmine rice is a long-grain variety, it contains more amylopectin, a highly branched starch molecule. Regular long-grain rice has more amylose. When amylose is digested, less glucose is freed at once than from the highly branched amylopectin chains, so high-amylose rice has a lower glucose response than high-amylopectin varieties (Mechanisms of the Effects of Grains on Insulin and Glucose Responses, Hallfrisch J, Behall KM, Journal of theAmerican College of Nutrition, June 2000, 19:320S-325S).

Those consuming a high-GI meal 4 hours before bedtime fell asleep more quickly (9.0 ± 6.2 minutes) than those eating a low-GI meal (17.5 ± 6.2 minutes). Eating the high-GI meal 4 hours before bedtime also put them to sleep faster than eating the same meal 1 hour before bedtime (14.6 ± 9.9 minutes). The researchers observed no effects on other sleep variables.