Apparently we are still not eating enough vegetables.
Semin Nephrol. 2006 Nov;26(6):447-53. Links
The evolution-informed optimal dietary potassium intake of human beings greatly exceeds current and recommended intakes.

* Sebastian A,
* Frassetto LA,
* Sellmeyer DE,
* Morris RC Jr.

Department of Medicine, Divisions of Nephrology and Endocrinology, and the Moffitt/Mt. Zion General Clinical Research Center, UCSF, San Francisco, CA.

An organism best fits the environment described by its genes, an environment that prevailed during the time period (millions of years) when evolution naturally selected the genes of its ancestors-those who survived to pass on their genes. When an organism’s current environment differs from its ancestral one, the environment’s mismatch with the organism’s genome may result in functional disadvantages for the organism. The genetically conditioned nutritional requirements of human beings established themselves over millions of years in which ancestral hominins, living as hunter-gatherers, ate a diet markedly different from that of agriculturally dependent contemporary human beings. In that context, we sought to quantify the ancestral-contemporary dietary difference with respect to the supply of one of the body’s major mineral nutrients: potassium. In 159 retrojected Stone Age diets, human potassium intake averaged 400 +/- 125 mEq/d, which exceeds current and recommended intakes by more than a factor of 4. We accounted for the transition to the relatively potassium-poor modern diet by the fact that the modern diet has substantially replaced Stone Age amounts of potassium-rich plant foods (especially fruits, leafy greens, vegetable fruits, roots, and tubers), with energy-dense nutrient-poor foods (separated fats, oils, refined sugars, and refined grains), and with potassium-poor energy-rich plant foods (especially cereal grains) introduced by agriculture (circa 10,000 years ago). Given the fundamental physiologic importance of potassium, such a large magnitude of change in potassium intake invites the consideration in human beings of whether the quantitative values of potassium-influenced physiologic phenomena (eg, blood pressure, insulin and aldosterone secretion rates, and intracellular pH) currently viewed as normal, in fact disaccord with genetically conditioned norms. We discuss the potential implications of our findings in respect to human health and disease.