Data A.1 on page 136 and the followup study described in Exercise A.4 imply that mice that eat when they should normally be sleeping gain significantly more weight than mice that eat at the usual time. This is consistent with other studies showing that nighttime eating by people can foster weight gain and prediabetic changes in glucose control. A biologist at the Salk Institute in California notes, however, ‘‘Unfortunately, there is not even a single study in humans looking at [how] temporal spreading of caloric intake [affects fat deposition or glucose control].’’ In other words, observational studies show a relationship between eating a substantial portion of daily calories at night and weight gain, but no experiments have been conducted that might allow us to deduce a cause and effect relationship. Describe carefully the design of a randomized experiment to test whether eating at night rather than during the day increases weight gain in humans.

Exercise A.4

A study described in Data A.1 on page 136 found that mice exposed to dim light at night gained substantially more weight than mice that had complete darkness at night, despite the fact that calorie intake and activity levels were the same for the two groups. Was it the light that caused the mice to gain weight or was it the fact that they ate a large percentage of their calories when mice would normally be sleeping? The same researchers conducted a second experiment to answer this question. The second experiment matched the first in every respect except a new variable was added: half of the mice could eat whenever they wanted while the other half could only eat at night (the normal time mice eat). The mice in dim light that could eat whenever they wanted grew fatter, matching the results in the first experiment almost exactly. However, the mice in dim light that could only eat at night stayed as slim and healthy as the mice in darkness.