This tidbit appeared in a letter to the editor of the Annals of Improbable Research, and should be treated with all due respect. The full letter:

Difficult to Swallow

Distinguished editors:

I have been testing the long-believed, little-examined theory that 
three meals a day are optimal for health and longevity.

My data, derived from longitudinal studies during the years 1964-
1990 of more than 20,000 health professionals throughout Italy, 
appear to indicate that there are several optimal numbers. Three 
meals a day indeed are indeed best, but seven meals a day also 
provide very good results, as do nine meals a day.

The data also appear to indicate that 37 meals a day are optimal. 
This strikes me on the face of it as being flat-out wrong, but the 
data are there, and I am at a loss as to how to explain them. 
Perhaps one of your readers can offer an explanation.

Patrice M. Arruda, Ph.D.
Foggia, Italy